/*********************************************************************
//  created:    2013/06/19 - 18:40
//  filename:   ObstacleDetectionComponent.cpp
//
//  author:     Danilo Alves de Lima and Students of SY27
//              Copyright Heudiasyc UMR UTC/CNRS 6599
// 
//  version:    $Id: $
//
//  purpose:   Obstacle detection from stereo image data
//			   
//
*********************************************************************/
//#include "GeneralDefinitions.h"

#include "ObstacleDetectionComponent.h"

#include <iostream>
#include <string>
#include "opencv2/calib3d/calib3d.hpp"
#include "opencv2/core/core.hpp"

#include "Pacpus/kernel/ComponentFactory.h"
#include "Pacpus/kernel/DbiteException.h"
#include "Pacpus/kernel/DbiteFileTypes.h"
#include "Pacpus/kernel/Log.h"

using namespace std;
using namespace pacpus;

DECLARE_STATIC_LOGGER("pacpus.base.ObstacleDetectionComponent");

// Construct the factory
static ComponentFactory<ObstacleDetectionComponent> sFactory("ObstacleDetectionComponent");

const int kMaxFilepathLength = 512; // TODO: should be same for all images

static const string ObstacleDetectionMemoryName_ref = "DisparityMap-ref";
static const string ObstacleDetectionMemoryName_dispin = "DisparityMap-disp";

static const string ObstacleDetectionMemoryName_mask1 = "ObstacleDetection-mask1";
static const string ObstacleDetectionMemoryName_mask2 = "ObstacleDetection-mask2";
static const string ObstacleDetectionMemoryName_dispout = "ObstacleDetection-disp";

static const string ObstacleDetectionMemoryName_dispMapNorm = "ObstacleDetection-dispMapNorm";
static const string ObstacleDetectionMemoryName_result = "ObstacleDetection-result";

//////////////////////////////////////////////////////////////////////////
/*	ComparePoints1
	Description:
		Compare if the point 1 if less than 2 by the criteria of the higher y e higher x
	Parameters:
		pt1 = point 1
		pt2 = point 2
*/
bool ComparePoints1( cv::Point pt1, cv::Point pt2)
{
	/*
		· Strict: pred(X, X) is always false.

		· Weak: If !pred(X, Y) && !pred(Y, X), X==Y.

		· Ordering: If pred(X, Y) && pred(Y, Z), then pred(X, Z).
	*/
	if(pt1.y > pt2.y)
	{
		return true;
	}
	else if((pt1.y == pt2.y)&&(pt1.x >= pt2.x))
	{
		return true;
	}
	else
	{
		return false;
	}
}

/// Constructor.
ObstacleDetectionComponent::ObstacleDetectionComponent(QString name)
	: ComponentBase(name)
{   
	LOG_TRACE(Q_FUNC_INFO);

	setRecording(0);

	ANG_VARIATION = 20.0;
	ANG_VARIATION2 = 7.0;

	this->cam_width = 1280;		// Image width
	this->cam_height = 960;		// Image height
	this->cam_channels = 3;
	this->showdebug = false;	// Show frame acquired

	// Size of the image data sizeof(char)*width*height*channels
	this->mMaxImageInputSize1 = sizeof(char) * this->cam_width * this->cam_height * this->cam_channels;

	// Input data
	this->shmem_ref = 0;			// Shared memory control access to the image data
	this->shmem_dispin = 0;			// Shared memory control access to the image data

	// Output data
	this->shmem_mask1 = 0;		// Shared memory control access to the image data (free space mask)
	this->shmem_mask2 = 0;		// Shared memory control access to the image data (obstacles mask)
	this->shmem_dispout = 0;	// Shared memory control access to the image data (disparity map)
}

//////////////////////////////////////////////////////////////////////////
/// Destructor.
ObstacleDetectionComponent::~ObstacleDetectionComponent()
{
	LOG_TRACE(Q_FUNC_INFO);

	if(this->shmem_ref)
		delete shmem_ref;

	this->shmem_ref = NULL;

	if(this->shmem_dispin)
		delete shmem_dispin;

	this->shmem_dispin = NULL;

	if(this->shmem_mask1)
		delete shmem_mask1;

	this->shmem_mask1 = NULL;

	if(this->shmem_mask2)
		delete shmem_mask2;

	this->shmem_mask2 = NULL;

	if(this->shmem_dispout)
		delete shmem_dispout;

	this->shmem_dispout = NULL;

	if(this->shmem_dispMapNormalized)
		delete shmem_dispMapNormalized;

	this->shmem_dispMapNormalized = NULL;

	if(this->shmem_result)
		delete shmem_result;

	this->shmem_result = NULL;
}

//////////////////////////////////////////////////////////////////////////
/// Called by the ComponentManager to start the component
void ObstacleDetectionComponent::startActivity()
{
	LOG_TRACE(Q_FUNC_INFO);

	this->mMaxImageInputSize1 = sizeof(unsigned char) * this->cam_width * this->cam_height * this->cam_channels;
	this->mMaxImageInputSize2 = sizeof(unsigned short) * this->cam_width * this->cam_height;
	this->mMaxImageOutputSize = sizeof(unsigned char) * this->cam_width * this->cam_height;

	this->ref_mem_size = sizeof(TimestampedStructImage) + this->mMaxImageInputSize1;
	this->dispin_mem_size = sizeof(TimestampedStructImage) + this->mMaxImageInputSize2;
	this->mask1_mem_size = sizeof(TimestampedStructImage) + this->mMaxImageOutputSize;
	this->mask2_mem_size = sizeof(TimestampedStructImage) + this->mMaxImageOutputSize;
	this->dispout_mem_size = sizeof(TimestampedStructImage) + this->mMaxImageInputSize2;

	// Allocate memory position for the maximum expected data
	this->ref_mem = malloc(this->ref_mem_size);
	this->dispin_mem = malloc(this->dispin_mem_size);
	this->mask1_mem = malloc(this->mask1_mem_size); 
	this->mask2_mem = malloc(this->mask2_mem_size);
	this->dispout_mem = malloc(this->dispout_mem_size);

	this->shmem_ref = new ShMem(ObstacleDetectionMemoryName_ref.c_str(), this->ref_mem_size);

	this->shmem_dispin = new ShMem(ObstacleDetectionMemoryName_dispin.c_str(), this->dispin_mem_size);

	this->shmem_mask1 = new ShMem(ObstacleDetectionMemoryName_mask1.c_str(), this->mask1_mem_size);

	this->shmem_mask2 = new ShMem(ObstacleDetectionMemoryName_mask2.c_str(), this->mask2_mem_size);

	this->shmem_dispout = new ShMem(ObstacleDetectionMemoryName_dispout.c_str(), this->dispout_mem_size);

	this->shmem_dispMapNormalized = new ShMem(ObstacleDetectionMemoryName_dispMapNorm.c_str(), (this->cam_width * this->cam_height));

	this->shmem_result = new ShMem(ObstacleDetectionMemoryName_result.c_str(), (this->cam_width * this->cam_height * 3));

	// Run thread
	THREAD_ALIVE = true;
	start();
}

//////////////////////////////////////////////////////////////////////////
/// Called by the ComponentManager to stop the component
void ObstacleDetectionComponent::stopActivity()
{
	LOG_TRACE(Q_FUNC_INFO);

	if(THREAD_ALIVE)
	{
		// Stop thread
		THREAD_ALIVE = false;

		while(is_running)
		{
			msleep(/*MS_DELAY*/10);
		}

		if(this->shmem_ref)
			delete shmem_ref;

		this->shmem_ref = NULL;

		if(this->shmem_dispin)
			delete shmem_dispin;

		this->shmem_dispin = NULL;

		if(this->shmem_mask1)
			delete shmem_mask1;

		this->shmem_mask1 = NULL;

		if(this->shmem_mask2)
			delete shmem_mask2;

		this->shmem_mask2 = NULL;

		if(this->shmem_dispout)
			delete shmem_dispout;

		this->shmem_dispout = NULL;

		if(this->shmem_dispMapNormalized)
			delete shmem_dispMapNormalized;

		this->shmem_dispMapNormalized = NULL;

		if(this->shmem_result)
			delete shmem_result;

		this->shmem_result = NULL;

		// Free the malloc memories
		free(this->ref_mem);
		free(this->dispin_mem);
		free(this->mask1_mem); 
		free(this->mask2_mem);
		free(this->dispout_mem);
	}

	LOG_INFO("stopped component '" << name() << "'");
}

//////////////////////////////////////////////////////////////////////////
/// Called by the ComponentManager to pass the XML parameters to the
/// component
ComponentBase::COMPONENT_CONFIGURATION ObstacleDetectionComponent::configureComponent(XmlComponentConfig config)
{
	LOG_TRACE(Q_FUNC_INFO);

	// Initialize with default values
	InitDefault();

	if (config.getProperty("recording") != QString::null)
		setRecording(config.getProperty("recording").toInt());

	if (config.getProperty("cam_width") != QString::null)
		this->cam_width = config.getProperty("cam_width").toInt();

	if (config.getProperty("cam_height") != QString::null)
		this->cam_height = config.getProperty("cam_height").toInt();

	if (config.getProperty("cam_channels") != QString::null)
		this->cam_channels = config.getProperty("cam_channels").toInt();

	if (config.getProperty("min_disp") != QString::null)
		this->min_disp = config.getProperty("min_disp").toInt();

	if (config.getProperty("max_disp") != QString::null)
		this->max_disp = config.getProperty("max_disp").toInt();

	if (config.getProperty("min_disp_norm") != QString::null)
		this->min_disp_norm = config.getProperty("min_disp_norm").toInt();

	if (config.getProperty("max_disp_norm") != QString::null)
		this->max_disp_norm = config.getProperty("max_disp_norm").toInt();

	if (config.getProperty("showdebug") != QString::null)
		this->showdebug = (bool)config.getProperty("showdebug").toInt();

	if (config.getProperty("destiny_roi_x") != QString::null)
		this->destiny_roi_x = config.getProperty("destiny_roi_x").toInt();

	if (config.getProperty("destiny_roi_y") != QString::null)
		this->destiny_roi_y = config.getProperty("destiny_roi_y").toInt();

	if (config.getProperty("destiny_roi_width") != QString::null)
		this->destiny_roi_width = config.getProperty("destiny_roi_width").toInt();

	if (config.getProperty("destiny_roi_height") != QString::null)
		this->destiny_roi_height = config.getProperty("destiny_roi_height").toInt();

	if( ((this->destiny_roi_height != this->cam_height)||(this->destiny_roi_width != this->cam_width))&&
		((this->destiny_roi_height <= this->cam_height)&&(this->destiny_roi_width <= this->cam_width)) )
		this->use_roi = true;

	LOG_INFO("configured component '" << name() << "'");
    return ComponentBase::CONFIGURED_OK;
}

/** 
* Initialize default values
*/
void ObstacleDetectionComponent::InitDefault()
{
	// Default
	this->cam_width = this->destiny_roi_width =  1280;		// Image width
	this->cam_height = this->destiny_roi_height = 960;		// Image height
	
	this->destiny_roi_x = this->destiny_roi_y = 0;		// Destiny image roi x and y
	
	this->use_roi = false;
	this->min_disp = 0;
	this->max_disp = 255;

	this->min_disp_norm = 0;		// Minimum disparity value to equalize the disp map 16
	this->max_disp_norm = 255;		// Maximum disparity value to equalize the disp map 16

	this->showdebug = false;							// Show frame acquired
}

// Thread loop
void ObstacleDetectionComponent::run()
{
	LOG_TRACE(Q_FUNC_INFO);

	this->is_running = true;
	
	
	if(this->CurrentReferenceFrame.cols != this->cam_width)
	{
		this->CurrentReferenceFrame = cv::Mat(cv::Size(this->cam_width , this->cam_height), CV_MAKETYPE(CV_8U, this->cam_channels));
	}

	// Create the image in which will be read the disparities
	if(this->CurrentDisparityMap16.cols != this->cam_width)
	{
		this->CurrentDisparityMap16 = cv::Mat( this->cam_height, this->cam_width, CV_16S, cv::Scalar(0) );
	}

	if(this->CurrentSurfaceMask.cols != this->cam_width)
	{
		this->CurrentSurfaceMask = cv::Mat(this->cam_height, this->cam_width, CV_MAKETYPE(CV_8U, 1), cv::Scalar(0) );
	}

	if(this->CurrentObstaclesMask.cols != this->cam_width)
	{
		this->CurrentObstaclesMask = cv::Mat(this->cam_height, this->cam_width, CV_MAKETYPE(CV_8U, 1), cv::Scalar(0) );
	}

	// Images for type convertion
	cv::Mat Disp_map = cv::Mat( this->CurrentDisparityMap16.rows, this->CurrentDisparityMap16.cols, CV_8UC1 );
	cv::Mat Disp_map16;
	
	// Keeps the last image timestamp;
	road_time_t last_reading = 0;

	// Initialize the output images header
	this->Mask1ImageHeader.image.width = this->cam_width;
	this->Mask1ImageHeader.image.height = this->cam_height;
	this->Mask1ImageHeader.image.channels = 1;
	this->Mask1ImageHeader.image.width_step = (size_t)(this->Mask1ImageHeader.image.height * this->Mask1ImageHeader.image.channels);
	this->Mask1ImageHeader.image.data_size = this->mMaxImageOutputSize;

	this->Mask2ImageHeader.image.width = this->cam_width;
	this->Mask2ImageHeader.image.height = this->cam_height;
	this->Mask2ImageHeader.image.channels = 1;
	this->Mask2ImageHeader.image.width_step = (size_t)(this->Mask2ImageHeader.image.height * this->Mask2ImageHeader.image.channels);
	this->Mask2ImageHeader.image.data_size = this->mMaxImageOutputSize;

	this->DispOutImageHeader.image.width = this->cam_width;
	this->DispOutImageHeader.image.height = this->cam_height;
	this->DispOutImageHeader.image.channels = 1;
	this->DispOutImageHeader.image.width_step = (size_t)(sizeof(unsigned short)*this->DispOutImageHeader.image.height * this->DispOutImageHeader.image.channels);
	this->DispOutImageHeader.image.data_size = this->mMaxImageInputSize2;

	// Time measurement
	road_time_t init_time = 0;

	while (THREAD_ALIVE) 
	{
		init_time = road_time();

		//LOG_INFO("Grab new image");
		// header + image
		this->shmem_ref->read(this->ref_mem, this->ref_mem_size);
		this->shmem_dispin->read(this->dispin_mem, this->dispin_mem_size);

		// Header
		memcpy( &this->RefImageHeader, this->ref_mem, sizeof(TimestampedStructImage)); 
		memcpy( &this->DispInImageHeader, this->dispin_mem, sizeof(TimestampedStructImage)); 
	
		// Check image header
		bool is_ok = false;
		if( (this->RefImageHeader.image.data_size == this->mMaxImageInputSize1) && (this->RefImageHeader.time != last_reading) &&
			(this->DispInImageHeader.image.data_size == this->mMaxImageInputSize2) && (this->DispInImageHeader.time == this->RefImageHeader.time) )
		{
			is_ok = true;
			last_reading = this->RefImageHeader.time;
			
			/*std::cout << "Expected image w:  " << ImageHeader.image.width << std::endl;
			std::cout << "Expected image h:  " << ImageHeader.image.height << std::endl;
			std::cout << "Expected image c:  " << ImageHeader.image.channels << std::endl;
			std::cout << "Expected image data:  " << ImageHeader.image.data_size << std::endl;
			std::cout << "Expected image size:  " << image_mem << std::endl;*/
		}
		/*else
		{
			LOG_ERROR("Error in the image data size!");
		}*/

		if(is_ok)
		{
			// Image data
			memcpy( (unsigned char*)(this->CurrentReferenceFrame.data), (unsigned char*)((TimestampedStructImage*)this->ref_mem + 1), this->RefImageHeader.image.data_size); 
			memcpy( (unsigned short*)(this->CurrentDisparityMap16.data), (unsigned short*)((TimestampedStructImage*)this->dispin_mem + 1), this->DispInImageHeader.image.data_size); 
		
			if(this->showdebug)
			{			
				cv::namedWindow( "ObstacleDetectionComponent - Current Reference Frame", CV_WINDOW_NORMAL );
				cv::imshow("ObstacleDetectionComponent - Current Reference Frame",this->CurrentReferenceFrame);
				cv::waitKey(1);
			}

			//============================================= Obstacle Detection ================================================

			cv::Mat v_disp_map, u_disp_map;

			cv::Mat roi_disp = (this->use_roi) ? this->CurrentDisparityMap16(cv::Rect(this->destiny_roi_x, this->destiny_roi_y, this->destiny_roi_width, this->destiny_roi_height)) :
				this->CurrentDisparityMap16;

			// U/V disparity maps calculation
			this->CalcUVDisparityMapNorm(roi_disp, v_disp_map, u_disp_map, Disp_map, this->min_disp_norm, this->max_disp_norm);

			/*
			// Real disparity information
			Disp_map16 = this->CurrentDisparityMap16/16;
					
			// Display it as a CV_8UC1 image
			Disp_map16.convertTo( Disp_map, CV_8UC1);//, double(255)/double(this->max_disp - this->min_disp));

			if(this->showdebug)
			{		
				cv::namedWindow( "ObstacleDetectionComponent - Current Disparity Map", CV_WINDOW_AUTOSIZE );
				cv::imshow("ObstacleDetectionComponent - Current Disparity Map", Disp_map);
				cv::waitKey(1);
			}

			// U/V disparity maps calculation
			std::pair<cv::Mat,cv::Mat> par_uv = this->CalcUVDisparityMap(Disp_map);
			cv::Mat v_disp_map = par_uv.second;

			cv::Mat u_disp_map = par_uv.first;
			*/

			if(this->showdebug)
			{			
				cv::namedWindow( "ObstacleDetectionComponent - Current V Disparity Map", CV_WINDOW_AUTOSIZE );
				cv::imshow("ObstacleDetectionComponent - Current V Disparity Map", v_disp_map);
				
				cv::namedWindow( "ObstacleDetectionComponent - Current U Disparity Map", CV_WINDOW_AUTOSIZE );
				cv::imshow("ObstacleDetectionComponent - Current U Disparity Map", u_disp_map);

				cv::namedWindow( "ObstacleDetectionComponent - Current Disparity Map Normalized", CV_WINDOW_AUTOSIZE );
				cv::imshow("ObstacleDetectionComponent - Current Disparity Map Normalized", Disp_map);
				cv::waitKey(1);
			}

			//memcpy(this->dispMapNorm_mem, &this->dispMapNormImageHeader, sizeof(TimestampedStructImage)); 
			//memcpy((void*)((TimestampedStructImage*)this->dispMapNorm_mem + 1), (void*)Disp_map.data, this->dispMapNormImageHeader.image.data_size); 
			this->shmem_dispMapNormalized->write((void*)Disp_map.data, (this->cam_width * this->destiny_roi_height));

			// Image to detect near obstacles
			cv::Mat v_disp_map2 = v_disp_map.clone();

			// Find the driveable surface
			//cv::Mat color_v_disp_map1 = this->FindSurface(v_disp_map, v_disp_map2);
			cv::Mat color_v_disp_map1 = this->FindSurface2(v_disp_map, v_disp_map2);

			// Find near obstacles
			//cv::Mat color_v_disp_map2 = this->FindNearObstacles(v_disp_map2, this->min_disp, this->max_disp);
			std::pair<cv::Mat, cv::Mat> color_uv_disp_map = this->FindNearObstaclesUV(v_disp_map2, u_disp_map, this->min_disp, this->max_disp);
			
			if(this->showdebug)
			{
				cv::namedWindow("ObstacleDetectionComponent - Mapa de Disparidade V + Free Space",CV_WINDOW_AUTOSIZE);
				cv::imshow("ObstacleDetectionComponent - Mapa de Disparidade V + Free Space", color_v_disp_map1);

				cv::namedWindow("ObstacleDetectionComponent - Mapa de Disparidade V + Obstacles",CV_WINDOW_AUTOSIZE);
				cv::imshow("ObstacleDetectionComponent - Mapa de Disparidade V + Obstacles", color_uv_disp_map.first);

				cv::namedWindow("ObstacleDetectionComponent - Mapa de Disparidade U + Obstacles",CV_WINDOW_AUTOSIZE);
				cv::imshow("ObstacleDetectionComponent - Mapa de Disparidade U + Obstacles", color_uv_disp_map.second);
				cv::waitKey(1);
			}

			// Remap the v-disparity point in the original image and create binary masks
			//std::pair<cv::Mat, cv::Mat> masks = this->MaskSurface2(Disp_map, color_v_disp_map1, color_uv_disp_map.first, this->min_disp, this->max_disp, 1);
			std::pair<cv::Mat, cv::Mat> masks = this->MaskSurface3(Disp_map, color_v_disp_map1, color_uv_disp_map.first, color_uv_disp_map.second, this->min_disp, this->max_disp, 1);

			/*if(this->showdebug)
			{
				cv::namedWindow("Mapa de Disparidade V + Mask1",CV_WINDOW_AUTOSIZE);
				cv::imshow("Mapa de Disparidade V + Mask1", masks.first*255);

				cv::namedWindow("Mapa de Disparidade V + Mask2",CV_WINDOW_AUTOSIZE);
				cv::imshow("Mapa de Disparidade V + Mask2", masks.second*255);
			}*/

			//---------------------- Remove commun information -------------------------
				
			masks.second = masks.second - masks.second.mul( masks.first);

			//---------------------------------------------------------------------------

			//------------------ Write the result in the shared memory ------------------
			
			if(this->use_roi)
			{
				masks.first.copyTo(this->CurrentSurfaceMask(cv::Rect(this->destiny_roi_x, this->destiny_roi_y, this->destiny_roi_width, this->destiny_roi_height)));
				masks.second.copyTo(this->CurrentObstaclesMask(cv::Rect(this->destiny_roi_x, this->destiny_roi_y, this->destiny_roi_width, this->destiny_roi_height)));
			}
			else
			{
				this->CurrentSurfaceMask = masks.first;
				this->CurrentObstaclesMask = masks.second;
			}

			//----------------------------- Mask 1 --------------------------------------
			// Complete timestamp header of the mask image 1
			this->Mask1ImageHeader.time = this->DispInImageHeader.time;
			this->Mask1ImageHeader.timerange = this->DispInImageHeader.timerange;

			// Copy images header and data to memory
			memcpy(this->mask1_mem, &this->Mask1ImageHeader, sizeof(TimestampedStructImage)); 
			memcpy((void*)((TimestampedStructImage*)this->mask1_mem + 1), (void*)this->CurrentSurfaceMask.data, this->Mask1ImageHeader.image.data_size); 
			this->shmem_mask1->write(this->mask1_mem, this->mask1_mem_size);

			//----------------------------- Mask 2 --------------------------------------
			// Complete timestamp header of the mask image 2
			this->Mask2ImageHeader.time = this->DispInImageHeader.time;
			this->Mask2ImageHeader.timerange = this->DispInImageHeader.timerange;

			// Copy images header and data to memory
			memcpy(this->mask2_mem, &this->Mask2ImageHeader, sizeof(TimestampedStructImage)); 
			memcpy((void*)((TimestampedStructImage*)this->mask2_mem + 1), (void*)this->CurrentObstaclesMask.data, this->Mask2ImageHeader.image.data_size); 
			this->shmem_mask2->write(this->mask2_mem, this->mask2_mem_size);

			//------------------------- Disparity map out -------------------------------
			// Complete timestamp header of the disp image out
			this->DispOutImageHeader.time = this->DispInImageHeader.time;
			this->DispOutImageHeader.timerange = this->DispInImageHeader.timerange;

			// Copy images header and data to memory
			memcpy(this->dispout_mem, &this->DispOutImageHeader, sizeof(TimestampedStructImage)); 
			memcpy((void*)((TimestampedStructImage*)this->dispout_mem + 1), (void*)this->CurrentDisparityMap16.data, this->DispOutImageHeader.image.data_size); 
			this->shmem_dispout->write(this->dispout_mem, this->dispout_mem_size);
			//---------------------------------------------------------------------------

			// ----------------- Apply the mask in the reference image ------------------
		
			std::vector<cv::Mat> channels(3);
			cv::Mat reference;
			if(this->cam_channels == 1)
			{
				cv::cvtColor((this->CurrentReferenceFrame(cv::Rect(this->destiny_roi_x, this->destiny_roi_y, this->destiny_roi_width, this->destiny_roi_height))).clone(), reference, CV_GRAY2BGR);
			}
			else
			{
				reference = (this->CurrentReferenceFrame(cv::Rect(this->destiny_roi_x, this->destiny_roi_y, this->destiny_roi_width, this->destiny_roi_height))).clone();
			}

			cv::split(reference, channels);
			
			masks.second = 1 - masks.second;
			channels[1] = masks.second.mul(channels[0]); // Activate the red color as obstacles
			channels[2] = masks.second.mul(channels[1]); // Activate the red color as obstacles

			//masks.second = masks.second - masks.first;
			//channels[0] = (1 - masks.second).mul(channels[0]); // Activate the yellow color for unclassified area
			
			cv::merge(channels, reference);

			this->shmem_result->write((void*)reference.data, (this->destiny_roi_width * this->destiny_roi_height * 3));

			if(this->showdebug)
			{
				cv::namedWindow("ObstacleDetectionComponent - Final Classification", CV_WINDOW_AUTOSIZE);
				cv::imshow("ObstacleDetectionComponent - Final Classification", reference);
			}
			//---------------------------------------------------------------------------

			//==================================================================================================================

			//std::cout << componentName.toStdString() << " cicle time: " << (road_time() - init_time)/1000000.0 << std::endl;
		}
		else
		{
			msleep(/*MS_DELAY*/10);
		}

		if(this->showdebug)
			cv::waitKey(1); // Give the system permission

		//std::cout << componentName.toStdString() << " cicle time: " << (road_time() - init_time)/1000000.0 << std::endl;
	}

	this->is_running = false;

	// Destroy the window frame
	if(this->showdebug)
		cvDestroyAllWindows();
}

// Function to calculate the U/V disparity map
std::pair<cv::Mat, cv::Mat> ObstacleDetectionComponent::CalcUVDisparityMap(cv::Mat disp_map)
{
	int l, c, pixel;  // local variables for row, line and pixel
	unsigned char intensity;	  // pixel intensity 

	unsigned char* ptr1; // row pointer for 8 bits image 
	//unsigned short* ptr2; // row pointer for 16 bits image

    // U disparity map iamge
	cv::Mat u_disp = cv::Mat::zeros(cv::Size(disp_map.cols, 256), CV_8UC1);

	// V disparity map image
	cv::Mat v_disp = cv::Mat::zeros(cv::Size(256, disp_map.rows), CV_8UC1);
    
    // run accross the image and add 1 to the respective U/V disparity column
	for (l = 0; l < disp_map.rows; ++l)
    {
		ptr1 = disp_map.ptr<unsigned char>(l);
	
		for (c = 0; c < disp_map.cols; ++c)
		{
			intensity = (unsigned char)ptr1[c];
			
			if( (intensity > this->min_disp)&&(intensity < this->max_disp))
			{				
				pixel =  intensity*u_disp.cols + c;
				u_disp.data[pixel] = (unsigned char)(u_disp.data[pixel] + 1);

				pixel =  l*v_disp.cols + intensity;
				v_disp.data[pixel] = (unsigned char)(v_disp.data[pixel] + 1);				
			}
        }   
	}

	return std::make_pair(u_disp, v_disp);
}

/*	CalcUVDisparityMapNorm
	Description:
		Function to calculate the U/V disparity map from a disp map normalized
	Parameters:
		disp_map16 = original disparity map 16
		disp_map_norm = resulted disparity map normalized
		min_d_norm = Minimum disparity value to equalize the disp map 16
		max_d_norm = Maximum disparity value to equalize the disp map 16
*/
void ObstacleDetectionComponent::CalcUVDisparityMapNorm(cv::Mat disp_map16, cv::Mat &v_disp_map, cv::Mat &u_disp_map, cv::Mat &disp_map_norm, int min_d_norm, int max_d_norm)
{
	int l, c, pixel;  // local variables for row, line and pixel
	unsigned char intensity;	// pixel intensity 
	int intensity_norm;			// pixel intensity 

	unsigned char* ptr1; // row pointer for 8 bits image 
	unsigned short* ptr2; // row pointer for 16 bits image

	// Disparity map image normalized
	disp_map_norm = cv::Mat::zeros(cv::Size(disp_map16.cols, disp_map16.rows), CV_8UC1);

    // U disparity map image
	u_disp_map = cv::Mat::zeros(cv::Size(disp_map16.cols, 256), CV_8UC1);

	// V disparity map image
	v_disp_map = cv::Mat::zeros(cv::Size(256, disp_map16.rows), CV_8UC1);
    
    // percorre a imagem original e soma 1 na coluna do mapa de disparidade V com a mesma
	// intensidade do pixel
	for (l = 0; l < disp_map16.rows; ++l)
    {
		ptr1 = disp_map_norm.ptr<unsigned char>(l);
		ptr2 = disp_map16.ptr<unsigned short>(l);

		for (c = 0; c < disp_map16.cols; ++c)
		{
			intensity = (unsigned char)(ptr2[c]/16);
			intensity_norm = (int)((float)((ptr2[c]/16.0f - (float)min_d_norm)*(255.0f)/((float)max_d_norm - (float)min_d_norm)) + 0.5f);
						
			if( (intensity > this->min_disp)&&(intensity < this->max_disp)&&(intensity_norm > 0)&&(intensity_norm < 255))
			{
				pixel =  intensity_norm*u_disp_map.cols + c;
				u_disp_map.data[pixel] = (unsigned char)(u_disp_map.data[pixel] + 1);

				pixel =  l*v_disp_map.cols + intensity_norm;
				v_disp_map.data[pixel] = (unsigned char)(v_disp_map.data[pixel] + 1);

				ptr1[c] = (unsigned char)intensity_norm;
			}
        }   
	}

	return;
}

// Function to find the free space surface from a V-disparity map
cv::Mat ObstacleDetectionComponent::FindSurface(cv::Mat &v_disp_map, cv::Mat &v_disp_map2)
{
	// Parameters of canny and hough transform
	int tshold1 = 154;
	int tshold2 = 48;
	int n_points = 48; //59
	int minLineLenght = 35; //40
	int maxLineGap = 12;
		
		
	// Binary image
	cv::Mat Img_bin = v_disp_map.clone();//cvCloneImage(v_disp_map);

	// Color V disparity map with red lines
	cv::Mat color_img = cv::Mat( cv::Size(v_disp_map.cols, v_disp_map.rows), CV_8UC3 );

	// Convert to color image
	cv::cvtColor(v_disp_map, color_img, CV_GRAY2BGR);

	cv::equalizeHist( Img_bin, Img_bin);

	if(this->showdebug)
	{
		// Janela de exibicao
		cv::namedWindow("ObstacleDetectionComponent - Equalized Image",CV_WINDOW_AUTOSIZE);
		cv::imshow("ObstacleDetectionComponent - Equalized Image", Img_bin);
	}

	cv::Canny(Img_bin, Img_bin, tshold1, tshold2, 3);
	
	// Closing
	//cv::dilate(Img_bin, Img_bin, cv::Mat(), cv::Point(-1,-1), 2 );
	//cv::erode(Img_bin, Img_bin, cv::Mat(), cv::Point(-1,-1), 1 );

	if(this->showdebug)
	{
		// Janela de exibicao
		cv::namedWindow("ObstacleDetectionComponent - Binary Image",CV_WINDOW_AUTOSIZE);
		cv::imshow("ObstacleDetectionComponent - Binary Image", Img_bin);
	}
		
    std::vector<cv::Vec4i> lines;			//vector for storing the lines found by HoughLine

	// Probabilistic Hough Transform
	cv::HoughLinesP( Img_bin, lines, 1, CV_PI/180, n_points, minLineLenght, maxLineGap ); 

	//=============================== Use the lines filter to remove invalid segments ============================ 

	std::vector<cv::Point> nova_lista = this->LinesFiltering(lines);
	
	if(!nova_lista.empty())
	{
		cv::Point pt_ant = *(nova_lista.begin());

		// Filter the mean angle
		for(std::vector<cv::Point>::iterator it = nova_lista.begin(); it != nova_lista.end(); ++it)
		{		
			cv::line( color_img, pt_ant, *it, CV_RGB(255,0,0), 8, 8 );

			pt_ant = *it;
		}
	}

	//============================================================================================================

	//======================== Remove invalid line segments by slope angle only ==================================
	//if (lines.size() != 0)
	//{
	//	cv::Point pt1, pt2;
	//	double theta;

	//	for(int i = 0; i < (int)lines.size();++i)
	//	{
	//		pt1.x = lines[i][0];//(CvPoint*)cvGetSeqElem(lines,i);
	//		pt1.y = lines[i][1];
	//		pt2.x = lines[i][2];
	//		pt2.y = lines[i][3];

	//		CheckPoints(pt1, pt2); //Verifica a ordem dos pontos
	//	
	//		theta = Inclination(pt1, pt2); //calcula a inclinacao da reta encontrada

	//		// Valor atual do angulo em graus
	//		theta = ((theta*360.0)/(2.0*CV_PI));

	//		// Verifica se a reta possui inclinacao para ser possivel plano
	//		if(theta> (90.0 + ANG_VARIATION))
	//		{

	//			//Desenha as retas em vermelho
	//			cv::line( color_img, pt1, pt2, CV_RGB(255,0,0), 4, 8 );
	//		}
	//	}		
	//}
	//==========================================================================================================

	std::vector<cv::Mat> channels(3);

	// Get the V-disparity without detected red lines
	cv::split(color_img, channels);
	v_disp_map2 = channels[0];
	
	// Janela de exibicao
	//cv::namedWindow("Mapa de Disparidade V + Hough",CV_WINDOW_AUTOSIZE);
	//cv::imshow("Mapa de Disparidade V + Hough", color_img);
	//cv::imshow("Mapa de Disparidade V + Hough", v_disp_map2);

	return color_img;
}

/*	FindSurface2
	Description:
		Function to find the free space surface from a V-disparity map, based on the frontal plane. 
		Return the V-dysparity map with the red line representing the free surface.
	Parameters:
		v_disp_map = Original V-disparity map
		v_disp_map2 = Orignal V-disparity map less the surface detected
*/
cv::Mat ObstacleDetectionComponent::FindSurface2(cv::Mat &v_disp_map, cv::Mat &v_disp_map2)
{	
	// Parameters
	int tshold = 40;	// Min threshold for the first max value
	int tshold2 = 35;	// Min threshold for the step variation
	int maxLineLenght = 25;
	int min_disp_value = 20;

	// Binary image
	cv::Mat Img_bin = v_disp_map.clone();
	cv::Mat Img_mask = v_disp_map.clone();
	
	// Color V disparity map with red lines
	cv::Mat color_img = cv::Mat( cv::Size(v_disp_map.cols, v_disp_map.rows), CV_8UC3 );

	// Convert to color image
	cv::cvtColor(v_disp_map, color_img, CV_GRAY2BGR);


	//================= Segment the most probable road surface region ===================================
	/*cv::threshold(Img_mask, Img_mask, tshold, 1, CV_THRESH_BINARY);
	cv::dilate(Img_mask, Img_mask, cv::Mat(), cv::Point(-1,-1), 2 );
	cv::erode(Img_mask, Img_mask, cv::Mat(), cv::Point(-1,-1), 1 );

	if(this->showdebug)
	{
		cv::imshow( "ObstacleDetectionComponent - Img_mask", Img_mask*255 );
	}
	
	Img_bin = Img_bin.mul(Img_mask);*/
	//===================================================================================================

	// Generate grad_x and grad_y
	cv::Mat grad, grad_H, grad_S, grad_x, grad_y;
	cv::Mat abs_grad_x, abs_grad_y;
		
	// Gradient X
	//Scharr( src_gray, grad_x, ddepth, 1, 0, scale, delta, BORDER_DEFAULT );
	//cv::Sobel( Img_bin, grad_x, CV_16S, 1, 0, 3, 1, 0, cv::BORDER_DEFAULT );
	//cv::convertScaleAbs( grad_x, abs_grad_x );
	
	// Gradient Y
	//Scharr( src_gray, grad_y, ddepth, 0, 1, scale, delta, BORDER_DEFAULT );
	cv::Sobel( Img_bin, grad_y, CV_16S, 0, 1, 3, 1, 0, cv::BORDER_DEFAULT );
	cv::convertScaleAbs( grad_y, abs_grad_y );
	
	// Total Gradient (approximate)
	//cv::addWeighted( abs_grad_x, 0.2, abs_grad_y, 0.8, 0, grad );

	//abs_grad_y = abs_grad_y*255.0;
	abs_grad_y.convertTo(Img_mask, CV_8U);
	
	cv::threshold(Img_mask, Img_mask, tshold2, 1, CV_THRESH_BINARY);
	cv::dilate(Img_mask, Img_mask, cv::Mat(), cv::Point(-1,-1), 2 );
	cv::erode(Img_mask, Img_mask, cv::Mat(), cv::Point(-1,-1), 1 );

	Img_bin = Img_bin.mul(Img_mask);

	//cv::equalizeHist( abs_grad_y, Img_bin);

	if(this->showdebug)
	{
		//cv::imshow( "ObstacleDetectionComponent - Sobel X", abs_grad_x );
		cv::imshow( "ObstacleDetectionComponent - Sobel Y", abs_grad_y );
		cv::imshow( "ObstacleDetectionComponent - Img_mask", Img_mask*255 );
		//cv::imshow( "ObstacleDetectionComponent - Sobel", grad );
		//cv::imshow( "ObstacleDetectionComponent - Equalized Image", Img_bin);
	}

	//============================== Mark the most significative points as free space ==============================
	
	int row_step_count, last_row_step_count;		// Keep the sum of the last valid pixels in the step
	int previous_col = 0;

	int left_limit, right_limit; // Auxiliary variables

	for(int row = Img_bin.rows - 1; row > 0; --row)
	{
		row_step_count = 0;
		last_row_step_count = 0;
		left_limit = -1;
		right_limit = -1;		

		unsigned char* ptr1 = Img_bin.ptr<unsigned char>(row);

		if(previous_col == 0)
		{			
			for(int col = 0; col < Img_bin.cols; ++col)
			{
				// Find max
				if(ptr1[previous_col] < ptr1[col])
				{
					previous_col = col;

					tshold2 = (int)(0.95*ptr1[previous_col]); // adjust the tshold
				}
			}
			
			if(ptr1[previous_col] < tshold)
			{
				previous_col = 0; // position
			}
		}
				
		if(previous_col != 0)
		{
			int left = (previous_col - maxLineLenght >= 0) ? previous_col-maxLineLenght : 0;
			int right = (previous_col + maxLineLenght < Img_bin.cols) ? previous_col + maxLineLenght : Img_bin.cols;

			int limits_found = 0;
			int l_col = previous_col;
			int r_col = previous_col;

			int left_count = 0;
			int right_count = 0;

			while(!limits_found)
			{
				//------------------------------- To the left ----------------------------------
				if(left_limit == -1)
				{ 
					if(l_col > left)
						--l_col;
					else
						left_limit = l_col + left_count;
				}				

				// Find max
				if(ptr1[previous_col] < ptr1[l_col])
				{
					previous_col = l_col;
					tshold2 = (int)(0.95*ptr1[previous_col]); // adjust the tshold
				}

				// Check if the pixel intensity is lower than tshold
				if(left_count < 5)
				{
					if(ptr1[l_col] < tshold2)
					{
						++left_count;
					}
					else
					{
						left_count = 0;
					}
				}
				else
				{
					if(left_limit == -1)
						left_limit = l_col + left_count;
				}
				//------------------------------------------------------------------------------

				//------------------------------- To the right ----------------------------------
				if(right_limit == -1)
				{ 
					if(r_col < right)
						++r_col;
					else
						right_limit = r_col - right_count;
				}				

				// Find max
				if(ptr1[previous_col] < ptr1[r_col])
				{
					previous_col = r_col;
					tshold2 = (int)(0.95*ptr1[previous_col]); // adjust the tshold
				}

				// Check if the pixel intensity is lower than tshold
				if(right_count < 5)
				{
					if(ptr1[r_col] < tshold2)
					{
						++right_count;
					}
					else
					{
						right_count = 0;
					}
				}
				else
				{
					if(right_limit == -1)
						right_limit = r_col - right_count;
				}
				//------------------------------------------------------------------------------

				if((left_limit != -1)&&(right_limit != -1))
				{
					limits_found = 1;
				}
			}

			if((left_limit != -1)&&(right_limit != -1)&&(ptr1[previous_col] > tshold2))
			{
				cv::line( color_img, cv::Point(left_limit, row), cv::Point(right_limit, row), CV_RGB(255,0,0), 1, 8 );
			}
		}
	}

	//==============================================================================================================
			
	std::vector<cv::Mat> channels(3);

	// Get the V-disparity without detected red lines
	cv::split(color_img, channels);
	v_disp_map2 = channels[0];
	
	// Janela de exibicao
	//cv::namedWindow("Mapa de Disparidade V + Hough",CV_WINDOW_AUTOSIZE);
	//cv::imshow("Mapa de Disparidade V + Hough", color_img);
	//cv::imshow("Mapa de Disparidade V + Hough", v_disp_map2);

	return color_img;
}

// Function to find the free space surface from a V-disparity map with mean average
cv::Mat ObstacleDetectionComponent::FindAverageSurface(cv::Mat &v_disp_map, cv::Mat &v_disp_map2)
{
	// Parameters of threshold and hough transform
	int tshold1 = 3;
	int n_points = 48; //59
	int minLineLenght = 35; //40
	int maxLineGap = 12;

	// Imagem atual em 32F
	cv::Mat v_disp_map_current32F = cv::Mat(cv::Size(v_disp_map.cols, v_disp_map.rows), CV_32F, 1);

	// Converte tipo de imagem
	v_disp_map.convertTo(v_disp_map_current32F, CV_32F);//cvConvertScale(v_disp_map, v_disp_map_current32F);

	//Mean disparity map for noise attenuation
	static cv::Mat v_disp_map_mean = v_disp_map_current32F.clone();

	// Imagem binária da diferenca
	cv::Mat ImgBinaria = v_disp_map.clone();//cvCloneImage(v_disp_map);

	// Color V disparity map with red lines
	cv::Mat color_img = cv::Mat( cv::Size(v_disp_map.cols, v_disp_map.rows), CV_8UC3 );

	// Convert to color image
	cv::cvtColor(v_disp_map, color_img, CV_GRAY2BGR);

	// Running Average
	cv::accumulateWeighted(v_disp_map_current32F, v_disp_map_mean, 0.20);//cvRunningAvg(v_disp_map_current32F, v_disp_map_mean, 0.20);

	// Convert scale 
	v_disp_map_mean.convertTo( ImgBinaria, CV_8U); //cvConvertScale(v_disp_map_mean, ImgBinaria, 1.0, 0.0);

	// Janela de exibicao
	//cv::namedWindow("Imagem Media Movel",CV_WINDOW_AUTOSIZE);
	//cv::imshow("Imagem Media Movel", ImgBinaria);
	
	cv::threshold(ImgBinaria, ImgBinaria, tshold1, 255,CV_THRESH_BINARY);

	// Janela de exibicao
	//cv::namedWindow("Mapa de Disparidade V + binarizacao",CV_WINDOW_AUTOSIZE);
	//cv::imshow("Mapa de Disparidade V + binarizacao", ImgBinaria);

	// create 3x3 matrix
	//static cv::Mat kernel_3x3 = (cv::Mat_<unsigned char>(3,3) << 1, 0, 0, 1, 1, 0, 1, 1, 1);
	
	// Fechamento
	//cv::erode(ImgBinaria,ImgBinaria,NULL, cv::Point(-1,-1), 1);
	//cv::erode(ImgBinaria,ImgBinaria,kernel_3x3, cv::Point(1,1), 1);

	// Janela de exibicao
	//cv::namedWindow("Mapa de Disparidade V + binarizacao + erode",CV_WINDOW_AUTOSIZE);
	//cv::imshow("Mapa de Disparidade V + binarizacao + erode", ImgBinaria);

	// Fechamento
	//cv::dilate(ImgBinaria,ImgBinaria,NULL, cv::Point(-1,-1), 1);
	//cv::erode(ImgBinaria,ImgBinaria,NULL, cv::Point(-1,-1), 2);

    std::vector<cv::Vec4i> lines;			//vector for storing the lines found by HoughLine

	// Probabilistic Hough Transform
	cv::HoughLinesP( ImgBinaria, lines, 1, CV_PI/180, n_points, minLineLenght, maxLineGap ); 

	//=============================== Use the lines filter to remove invalid segments ============================ 

	//std::vector<cv::Point> nova_lista = this->LinesFiltering(lines);
	//
	//if(!nova_lista.empty())
	//{
	//	cv::Point pt_ant = *(nova_lista.begin());

	//	// Filter the mean angle
	//	for(std::vector<cv::Point>::iterator it = nova_lista.begin(); it != nova_lista.end(); ++it)
	//	{		
	//		cv::line( color_img, pt_ant, *it, CV_RGB(255,0,0), 3, 8 );

	//		pt_ant = *it;
	//	}
	//}

	//============================================================================================================

	//======================== Remove invalid line segments by slope angle only ==================================
	if (lines.size() != 0)
	{
		cv::Point pt1, pt2;
		double theta;

		for(int i = 0; i < (int)lines.size();++i)
		{
			pt1.x = lines[i][0];//(CvPoint*)cvGetSeqElem(lines,i);
			pt1.y = lines[i][1];
			pt2.x = lines[i][2];
			pt2.y = lines[i][3];

			CheckPoints(pt1, pt2); //Verifica a ordem dos pontos
		
			theta = Inclination(pt1, pt2); //calcula a inclinacao da reta encontrada

			// Valor atual do angulo em graus
			theta = ((theta*360.0)/(2.0*CV_PI));

			// Verifica se a reta possui inclinacao para ser possivel plano
			if(theta> (90.0 + ANG_VARIATION))
			{

				//Desenha as retas em vermelho
				cv::line( color_img, pt1, pt2, CV_RGB(255,0,0), 6, 8 );
			}
		}		
	}
	//==========================================================================================================

	std::vector<cv::Mat> channels(3);

	// Get the V-disparity without detected red lines
	cv::split(color_img, channels);
	v_disp_map2 = channels[0];
	
	// Janela de exibicao
	//cv::namedWindow("Mapa de Disparidade V + Hough",CV_WINDOW_AUTOSIZE);
	//cv::imshow("Mapa de Disparidade V + Hough", color_img);
	//cv::imshow("Mapa de Disparidade V + Hough", v_disp_map2);

	return color_img;
}

// Function to find the near obstacles from a v-Disparity map
cv::Mat ObstacleDetectionComponent::FindNearObstacles(cv::Mat v_disp_map, int min_d, int max_d)
{
	// Parameters of threshold and hough transform
	int tshold1 = 3;
	int n_points = 15;
	int minLineLenght = 10;
	int maxLineGap = 12;

	// Image to be processed
	cv::Mat v_disp_map_aux = v_disp_map.clone();

	// Image with the obstacles highlighted in red
	cv::Mat color_img = cv::Mat( cv::Size(v_disp_map.cols, v_disp_map.rows), CV_8UC3);

	// Convert color space
	cv::cvtColor(v_disp_map_aux, color_img, CV_GRAY2BGR);

	// Image binarization
	cv::threshold(v_disp_map_aux, v_disp_map_aux, tshold1, 255, CV_THRESH_BINARY);

	// Janela de exibicao
	//cv::namedWindow("Mapa de Disparidade V + Threshold 2",CV_WINDOW_AUTOSIZE);
	//cv::imshow("Mapa de Disparidade V + Threshold 2", v_disp_map_aux);

	// Define valid ROI
	cv::Mat v_disp_map_aux_ROI = v_disp_map_aux(cv::Rect(min_d, 0, max_d-min_d+1, v_disp_map_aux.rows));
	cv::Mat color_img_ROI = color_img( cv::Rect(min_d, 0, max_d-min_d+1, v_disp_map_aux.rows));

	std::vector<cv::Vec4i> lines;			//vector for storing the lines found by HoughLine

	// Probabilistic hough transform
	cv::HoughLinesP( v_disp_map_aux_ROI, lines, 1, CV_PI/180, n_points, minLineLenght, maxLineGap ); 
			
	if (lines.size() != 0)
	{
		cv::Point pt1, pt2;
		double theta;

		for(int i = 0; i < (int)lines.size(); ++i)
		{
			pt1.x = lines[i][0];//(CvPoint*)cvGetSeqElem(lines,i);
			pt1.y = lines[i][1];
			pt2.x = lines[i][2];
			pt2.y = lines[i][3];

			this->CheckPoints(pt1, pt2); //Verify the points order
		
			theta = this->Inclination(pt1, pt2); // line slope

			// In degrees
			theta = ((theta*360.0)/(2.0*CV_PI));

			// Verifica se a reta possui inclinacao para ser possivel plano
			if((theta < (90.0 + ANG_VARIATION2))&& (theta > (90.0 - ANG_VARIATION2)))
			{								
				//Desenha as retas em vermelho
				cv::line( color_img_ROI, pt1, pt2, CV_RGB(255,0,0), 5, 8 );
			}
		}		
	}
		
	// Janela de exibicao
	//cv::namedWindow("Mapa de Disparidade V + Hough1",CV_WINDOW_AUTOSIZE);
	//cv::imshow("Mapa de Disparidade V + Hough1", v_disp_map_aux);

	// Janela de exibicao
	//cv::namedWindow("Mapa de Disparidade V + Hough2",CV_WINDOW_AUTOSIZE);
	//cv::imshow("Mapa de Disparidade V + Hough2", color_img);
	
	return color_img;
}

// Function to find the near obstacles from the v/u-Disparity maps
std::pair<cv::Mat, cv::Mat> ObstacleDetectionComponent::FindNearObstaclesUV(cv::Mat v_disp_map, cv::Mat u_disp_map, int min_d, int max_d)
{
	// Parameters of threshold and hough transform
	int tshold1 = 3;
	int tshold2 = 15;
	int n_points = 15;
	int minLineLenght = 10;
	int maxLineGap = 12;

	std::vector<cv::Mat> channels(3);

	//=================================== V-disparity map process =============================================================
	//// Image to be processed
	//cv::Mat v_disp_map_aux = v_disp_map.clone();
	//	
	//// Image with the obstacles highlighted in red
	//cv::Mat v_color_img = cv::Mat( cv::Size(v_disp_map.cols, v_disp_map.rows), CV_8UC3);

	//// Convert color space
	//cv::cvtColor(v_disp_map_aux, v_color_img, CV_GRAY2BGR);

	//cv::equalizeHist( v_disp_map_aux, v_disp_map_aux);

	//// Image binarization
	//cv::threshold(v_disp_map_aux, v_disp_map_aux, tshold1, 255, CV_THRESH_BINARY);

	//// Janela de exibicao
	//if(this->showdebug)
	//{
	//	cv::namedWindow("Mapa de Disparidade V + Threshold 2",CV_WINDOW_AUTOSIZE);
	//	cv::imshow("Mapa de Disparidade V + Threshold 2", v_disp_map_aux);
	//}

	//// Define valid ROI
	//cv::Mat v_disp_map_aux_ROI = v_disp_map_aux(cv::Rect(min_d, 0, max_d-min_d+1, v_disp_map_aux.rows));
	//cv::Mat v_color_img_ROI = v_color_img( cv::Rect(min_d, 0, max_d-min_d+1, v_disp_map_aux.rows));

	//std::vector<cv::Vec4i> lines;			//vector for storing the lines found by HoughLine

	//// Probabilistic hough transform
	//cv::HoughLinesP( v_disp_map_aux_ROI, lines, 1, CV_PI/180, n_points, minLineLenght, maxLineGap ); 
	//		
	//if (lines.size() != 0)
	//{
	//	cv::Point pt1, pt2;
	//	double theta;

	//	for(int i = 0; i < (int)lines.size(); ++i)
	//	{
	//		pt1.x = lines[i][0];//(CvPoint*)cvGetSeqElem(lines,i);
	//		pt1.y = lines[i][1];
	//		pt2.x = lines[i][2];
	//		pt2.y = lines[i][3];

	//		this->CheckPoints(pt1, pt2); //Verify the points order
	//	
	//		theta = this->Inclination(pt1, pt2); // line slope

	//		// In degrees
	//		theta = ((theta*360.0)/(2.0*CV_PI));

	//		// Verifica se a reta possui inclinacao para ser possivel plano
	//		//if((theta < (90.0 + ANG_VARIATION2))&& (theta > (90.0 - ANG_VARIATION2)))
	//		//{								
	//			//Desenha as retas em vermelho
	//			cv::line( v_color_img_ROI, pt1, pt2, CV_RGB(255,0,0), 5, 8 );
	//		//}
	//	}		
	//}
	//	
	//// Janela de exibicao
	////cv::namedWindow("Mapa de Disparidade V + Hough1",CV_WINDOW_AUTOSIZE);
	////cv::imshow("Mapa de Disparidade V + Hough1", v_disp_map_aux);

	//// Janela de exibicao
	////cv::namedWindow("Mapa de Disparidade V + Hough2",CV_WINDOW_AUTOSIZE);
	////cv::imshow("Mapa de Disparidade V + Hough2", color_img);

	//================================================================================================================================

	//============================================== V-disparity map process =========================================================
	// Image to be processed
	cv::Mat v_disp_map_aux = v_disp_map.clone();
		
	// Image with the obstacles highlighted in red
	cv::Mat v_color_img = cv::Mat( v_disp_map.rows, v_disp_map.cols, CV_8UC3, cv::Scalar(0,0,0));
	
	// Convert color space
	//cv::cvtColor(v_disp_map_aux, v_color_img, CV_GRAY2BGR);

	//cv::equalizeHist( v_disp_map_aux, v_disp_map_aux);

	// Image binarization
	cv::threshold(v_disp_map_aux, v_disp_map_aux, tshold1, 255, CV_THRESH_BINARY);

	// Janela de exibicao
	if(this->showdebug)
	{
		cv::namedWindow("Mapa de Disparidade V + Threshold 2",CV_WINDOW_AUTOSIZE);
		cv::imshow("Mapa de Disparidade V + Threshold 2", v_disp_map_aux);
	}

	// Define valid ROI
	cv::Mat v_disp_map_aux_ROI = v_disp_map_aux(cv::Rect(min_d, 0, max_d-min_d+1, v_disp_map_aux.rows));
	cv::Mat v_color_img_ROI = v_color_img( cv::Rect(min_d, 0, max_d-min_d+1, v_disp_map_aux.rows));

	cv::split( v_color_img, channels);
	channels[2] = v_disp_map_aux; 
	cv::merge( channels, v_color_img);

	//================================================================================================================================

	//=============================================== U-disparity map process ========================================================
	// Image to be processed
	cv::Mat u_disp_map_aux = u_disp_map.clone();
	
	// Image with the obstacles highlighted in red
	cv::Mat u_color_img = cv::Mat( cv::Size(u_disp_map.cols, u_disp_map.rows), CV_8UC3, cv::Scalar(0, 0, 0));

	// Convert color space
	//cv::cvtColor(u_disp_map_aux, u_color_img, CV_GRAY2BGR);

	// Image binarization
	cv::threshold(u_disp_map_aux, u_disp_map_aux, tshold2, 255, CV_THRESH_BINARY);
		
	// Closing
	cv::dilate(u_disp_map_aux, u_disp_map_aux, cv::Mat(), cv::Point(-1,-1), 1 );
	cv::erode(u_disp_map_aux, u_disp_map_aux, cv::Mat(), cv::Point(-1,-1), 1 );
	
	
	if(this->showdebug)
	{
		// Janela de exibicao
		cv::namedWindow("ObstacleDetectionComponent - Image bin u-disp",CV_WINDOW_AUTOSIZE);
		cv::imshow("ObstacleDetectionComponent - Image bin u-disp", u_disp_map_aux);
	}
	
	cv::split( u_color_img, channels);
	channels[2] = u_disp_map_aux; 
	cv::merge( channels, u_color_img);

	// Define valid ROI
	//cv::Mat u_disp_map_aux_ROI = v_disp_map_aux(cv::Rect(min_d, 0, max_d-min_d+1, v_disp_map_aux.rows));
	//cv::Mat u_color_img_ROI = v_color_img( cv::Rect(min_d, 0, max_d-min_d+1, v_disp_map_aux.rows));
	//================================================================================================================================

	return std::make_pair(v_color_img, u_color_img);
}

/*	LinesFiltering
	Description:
		Filter the detected lines related to the distance and angle between them.
	Parameters:
		lines = line list (point 1 and 2)
*/
std::vector<cv::Point> ObstacleDetectionComponent::LinesFiltering(std::vector<cv::Vec4i> lines)
{
	std::vector<cv::Point> lista_pontos;

	std::vector<cv::Point> lista_retorno;

	std::vector<double> angles;

	cv::Point pt1, pt2;
	double theta;
	double angulo_medio = 0.0;

	if (lines.size() != 0)
	{
		//std::cout << "Found "<< lines.size() << " lines!\n";

		// Filtro de angulo maximo
		for(int i = 0; i < (int)lines.size(); ++i)
		{
			pt1.x = lines[i][0];//(CvPoint*)cvGetSeqElem(lines,i);
			pt1.y = lines[i][1];
			pt2.x = lines[i][2];
			pt2.y = lines[i][3];

			CheckPoints(pt1, pt2); //Verifica a ordem dos pontos
		
			theta = Inclination(pt1, pt2); //calcula a inclinacao da reta encontrada

			// Valor atual do angulo em graus
			theta = ((theta*360.0)/(2.0*CV_PI));

			//std::cout << "Angle "<< theta << "!\n";

			// Verifica se a reta possui inclinacao para ser possivel plano
			if(theta> (90.0 + ANG_VARIATION))
			{
				lista_pontos.push_back(pt1);
				lista_pontos.push_back(pt2);
				
				angles.push_back(theta);
			}
		}

		if(!lista_pontos.empty())
		{
			std::sort(lista_pontos.begin(), lista_pontos.end(), ComparePoints1);
			angulo_medio = this->CalcMedian(angles);
		}

		CvPoint last_item;

		// Percorre os pontos ordenados e procura a casca convexa mais a esquerda
		for(std::vector<cv::Point>::iterator it = lista_pontos.begin(); it != lista_pontos.end(); ++it)
		{
			if(lista_retorno.empty())
			{
				lista_retorno.push_back(*it);

				last_item = *it;
			}
			else
			{
				if( (it + 1) != lista_pontos.end())
				{
					// Se tiverem o mesmo y, substitui o ponto
					if(last_item.y == (*it).y)
					{
						// Troca o ultimo elemento
						lista_retorno.pop_back();
												
						lista_retorno.push_back(*it);

						last_item = *it;
					}
					else
					{
						theta = Inclination(*it, last_item); //calcula a inclinacao da reta encontrada

						// Valor atual do angulo em graus
						theta = ((theta*360.0)/(2.0*CV_PI));

						lista_retorno.push_back(*it);

						// Percorre os pontos ordenados e procura o maior angulo entre a casca convexa mais a esquerda
						for(std::vector<cv::Point>::iterator it2 = it + 1; it2 != lista_pontos.end(); ++it2)
						{
							// Verifica se o angulo atual e menor que o anterior
							if(theta < ((Inclination(*it2, last_item)*360.0)/(2.0*CV_PI)))
							{
								// Troca o ultimo elemento
								lista_retorno.pop_back();
												
								lista_retorno.push_back(*it2);

								theta = Inclination(*it2, last_item); //calcula a inclinacao da reta encontrada

								// Valor atual do angulo em graus
								theta = ((theta*360.0)/(2.0*CV_PI));

								it = it2;
							}
						} // for

						last_item = *it;
					} // else
				} // if( (it + 1) != lista_pontos.end())
			} // else
		} // for

		if(!lista_retorno.empty())
		{
			last_item = *(lista_retorno.begin());

			// Filtro de angulo medio
			for(std::vector<cv::Point>::iterator it = lista_retorno.begin() + 1; it != lista_retorno.end(); ++it)
			{		
				theta = Inclination(*it, last_item); //calcula a inclinacao da reta encontrada

				// Valor atual do angulo em graus
				theta = ((theta*360.0)/(2.0*CV_PI));

				// Verifica se a reta possui inclinacao fora da media
				if(theta < (0.95*angulo_medio))//((theta > (1.05*angulo_medio))&&(theta < (0.95*angulo_medio)))
				{
					lista_retorno.resize(it - lista_retorno.begin());

					break;
				}

				last_item = *it;
			}
		}

	}

	return lista_retorno;
}

//Function to check the points order, making the second one with the highest y ever
void ObstacleDetectionComponent::CheckPoints(cv::Point &pt1, cv::Point &pt2)
{
	int aux_x, aux_y;
	
	if(pt1.y > pt2.y)
	{
		aux_x = pt1.x;
		aux_y = pt1.y;

		pt1.x = pt2.x;
		pt1.y = pt2.y;

		pt2.x = aux_x; 
		pt2.y = aux_y;
	}

	return;
}

// Function to calculate the line slope of pt1 to pt2
double ObstacleDetectionComponent::Inclination(cv::Point pt1, cv::Point pt2)
{
	double theta; //angle
		
	theta = fabs((atan2((pt1.y-pt2.y+0.0),(pt1.x-pt2.x+0.0)))); // slope

	return theta;
}

/*	CalcMedian
	Description:
		Calcule the median value of a vector.
	Parametros:
		vector = vector with data to calculate the median
*/
template<class A>
A ObstacleDetectionComponent::CalcMedian(std::vector<A> vetor) const
{
	A mediana;

	std::sort(vetor.begin(), vetor.end());

	mediana = vetor[(int)((double)(vetor.size())/2.0 + 0.5) - 1];

	return mediana;
}

// Function to calculate the free space (v_disp_1) and obstacles (v_disp_2) masks from 
// a red highlighted V-Disparity map
std::pair<cv::Mat, cv::Mat> ObstacleDetectionComponent::MaskSurface2(cv::Mat disp_map, cv::Mat v_disp_1, cv::Mat v_disp_2, int min_d, int max_d, int value)
{
	// Free space mask
	cv::Mat mask1;

	// Obstacle Mask
	cv::Mat mask2;
	
	// Fill the destiny images with background value
	if(value == 1)
	{
		mask1 = cv::Mat::zeros( cv::Size(disp_map.cols, disp_map.rows), CV_8UC1 );
		mask2 = cv::Mat::zeros( cv::Size(disp_map.cols, disp_map.rows), CV_8UC1 );
	}
	else
	{
		mask1 = cv::Mat::ones( cv::Size(disp_map.cols, disp_map.rows), CV_8UC1 );
		mask2 = cv::Mat::ones( cv::Size(disp_map.cols, disp_map.rows), CV_8UC1 );
	}

	int l, c, pixel, v_disp_pixel;	// local variables for row, column and pixel
	unsigned char intensity;		// pixel intensity
	unsigned char intensity_B, intensity_G, intensity_R;	//pixel intensity

	/*uint8_t* pixelPtr_mask1 = (uint8_t*)mask1.data;
	uint8_t* pixelPtr_mask2 = (uint8_t*)mask2.data;
	uint8_t* pixelPtr_disp_map = (uint8_t*)disp_map.data;
	uint8_t* pixelPtr_v_disp1 = (uint8_t*)v_disp_1.data;
	uint8_t* pixelPtr_v_disp2 = (uint8_t*)v_disp_2.data;*/
	
    // run the images associating the red pixels in the v-disparities in the mask1 e mask2
	for (l = 0; l < disp_map.rows; ++l)
    {
		for (c = 0; c < disp_map.cols; ++c)
        {
			pixel = l*disp_map.cols + c;

			intensity = (unsigned char)disp_map.data[pixel];

			//--------- Marca os planos trafegaveis --------------------
			// Pixel Azul
			v_disp_pixel = l*v_disp_1.cols*3 + intensity*3;
			intensity_B = (unsigned char)v_disp_1.data[v_disp_pixel];

			// Pixel Verde
			v_disp_pixel = l*v_disp_1.cols*3 + intensity*3 + 1;
			intensity_G = (unsigned char)v_disp_1.data[v_disp_pixel];

			// Pixel Vermelho
			v_disp_pixel = l*v_disp_1.cols*3 + intensity*3 + 2;
			intensity_R = (unsigned char)v_disp_1.data[v_disp_pixel];
			
			if((intensity_B == 0)&&(intensity_G == 0)&&(intensity_R == 255))
			{
				mask1.data[pixel] = value;	
			}
			//----------------------------------------------------------

			//--------- Marca os obstaculos ----------------------------					
			if( (intensity >= min_d)&&(intensity <= max_d))
			{				
				// Pixel Azul
				v_disp_pixel = l*v_disp_2.cols*3 + intensity*3;
				intensity_B = (unsigned char)v_disp_2.data[v_disp_pixel];

				// Pixel Verde
				v_disp_pixel = l*v_disp_2.cols*3 + intensity*3 + 1;
				intensity_G = (unsigned char)v_disp_2.data[v_disp_pixel];

				// Pixel Vermelho
				v_disp_pixel = l*v_disp_2.cols*3 + intensity*3 + 2;
				intensity_R = (unsigned char)v_disp_2.data[v_disp_pixel];
				
				if((intensity_B == 0)&&(intensity_G == 0)&&(intensity_R == 255))
				{
					mask2.data[pixel] = value;
				}
			}
			//----------------------------------------------------------
        }   
	}

	/*cv::Mat element = cv::getStructuringElement( cv::MORPH_RECT,
                                       cv::Size( 3, 3 ),
                                       cv::Point( 1, 1 ) );*/

	// Espande as regioes brancas da imagem
	if(value == 1)
	{
		//cv::dilate(mask1, mask1, element, cv::Point(1,1), 2);
		//cv::dilate(mask2, mask2, element, cv::Point(1,1), 2);
		
		cv::dilate(mask1, mask1, cv::Mat(), cv::Point(-1,-1), 2 );
		cv::dilate(mask2, mask2, cv::Mat(), cv::Point(-1,-1), 2 );
	}
	else
	{
		//cv::erode(mask1, mask1, element, cv::Point(1,1), 2);
		//cv::erode(mask2, mask2, element, cv::Point(1,1), 2);

		cv::erode(mask1,mask1,NULL, cv::Point(-1,-1), 2);
		cv::erode(mask2,mask2,NULL, cv::Point(-1,-1), 2);
	}

	//cvNamedWindow("Imagem da mascara1",CV_WINDOW_AUTOSIZE);
	//cvShowImage("Imagem da mascara1", mask1);

	//cvNamedWindow("Imagem da mascara2",CV_WINDOW_AUTOSIZE);
	//cvShowImage("Imagem da mascara2", mask2);

	return std::make_pair(mask1, mask2);
}

// Function to calculate the free space (v_disp_1) and obstacles (v_disp_2/u_disp) masks from 
// a red highlighted V-Disparity map
std::pair<cv::Mat, cv::Mat> ObstacleDetectionComponent::MaskSurface3(cv::Mat disp_map, cv::Mat v_disp_1, cv::Mat v_disp_2, cv::Mat u_disp, int min_d, int max_d, int value)
{
	// Free space mask
	cv::Mat mask1;

	// Obstacle Mask
	cv::Mat mask2;
	
	// Fill the destiny images with background value
	if(value == 1)
	{
		mask1 = cv::Mat::zeros( cv::Size(disp_map.cols, disp_map.rows), CV_8UC1 );
		mask2 = cv::Mat::zeros( cv::Size(disp_map.cols, disp_map.rows), CV_8UC1 );
	}
	else
	{
		mask1 = cv::Mat::ones( cv::Size(disp_map.cols, disp_map.rows), CV_8UC1 );
		mask2 = cv::Mat::ones( cv::Size(disp_map.cols, disp_map.rows), CV_8UC1 );
	}

	int l, c, pixel, v_disp_pixel, u_disp_pixel;	// local variables for row, column and pixel
	unsigned char intensity;		// pixel intensity
	unsigned char intensity_B_v, intensity_G_v, intensity_R_v;	//pixel intensity
	unsigned char intensity_B_u, intensity_G_u, intensity_R_u;	//pixel intensity

	/*uint8_t* pixelPtr_mask1 = (uint8_t*)mask1.data;
	uint8_t* pixelPtr_mask2 = (uint8_t*)mask2.data;
	uint8_t* pixelPtr_disp_map = (uint8_t*)disp_map.data;
	uint8_t* pixelPtr_v_disp1 = (uint8_t*)v_disp_1.data;
	uint8_t* pixelPtr_v_disp2 = (uint8_t*)v_disp_2.data;*/
	
    // run the images associating the red pixels in the v-disparities in the mask1 e mask2
	for (l = 0; l < disp_map.rows; ++l)
    {
		for (c = 0; c < disp_map.cols; ++c)
        {
			pixel = l*disp_map.cols + c;

			intensity = (unsigned char)disp_map.data[pixel];

			//--------- Marca os planos trafegaveis --------------------
			// Pixel Azul
			v_disp_pixel = l*v_disp_1.cols*3 + intensity*3;
			intensity_B_v = (unsigned char)v_disp_1.data[v_disp_pixel];

			// Pixel Verde
			v_disp_pixel = l*v_disp_1.cols*3 + intensity*3 + 1;
			intensity_G_v = (unsigned char)v_disp_1.data[v_disp_pixel];

			// Pixel Vermelho
			v_disp_pixel = l*v_disp_1.cols*3 + intensity*3 + 2;
			intensity_R_v = (unsigned char)v_disp_1.data[v_disp_pixel];
			
			if((intensity_B_v == 0)&&(intensity_G_v == 0)&&(intensity_R_v == 255))
			{
				mask1.data[pixel] = value;	
			}
			//----------------------------------------------------------

			//--------- Marca os obstaculos ----------------------------					
			if( (intensity >= min_d)&&(intensity <= max_d))
			{				
				// Pixel Azul
				v_disp_pixel = l*v_disp_2.cols*3 + intensity*3;
				u_disp_pixel = intensity*u_disp.cols*3 + c*3;
				intensity_B_v = (unsigned char)v_disp_2.data[v_disp_pixel];
				intensity_B_u = (unsigned char)u_disp.data[u_disp_pixel];

				// Pixel Verde
				v_disp_pixel = l*v_disp_2.cols*3 + intensity*3 + 1;
				u_disp_pixel = intensity*u_disp.cols*3 + c*3 + 1;
				intensity_G_v = (unsigned char)v_disp_2.data[v_disp_pixel];
				intensity_G_u = (unsigned char)u_disp.data[u_disp_pixel];

				// Pixel Vermelho
				v_disp_pixel = l*v_disp_2.cols*3 + intensity*3 + 2;
				u_disp_pixel = intensity*u_disp.cols*3 + c*3 + 2;
				intensity_R_v = (unsigned char)v_disp_2.data[v_disp_pixel];
				intensity_R_u = (unsigned char)u_disp.data[u_disp_pixel];
				
				if((intensity_B_v == 0)&&(intensity_G_v == 0)&&(intensity_R_v == 255)&&(intensity_B_u == 0)&&(intensity_G_u == 0)&&(intensity_R_u == 255))
				{
					mask2.data[pixel] = value;
				}
			}
			//----------------------------------------------------------
        }   
	}

	/*cv::Mat element = cv::getStructuringElement( cv::MORPH_RECT,
                                       cv::Size( 3, 3 ),
                                       cv::Point( 1, 1 ) );*/

	// Espande as regioes brancas da imagem
	if(value == 1)
	{
		//cv::dilate(mask1, mask1, element, cv::Point(1,1), 2);
		//cv::dilate(mask2, mask2, element, cv::Point(1,1), 2);
		
		cv::dilate(mask1, mask1, cv::Mat(), cv::Point(-1,-1), 2 );
		cv::dilate(mask2, mask2, cv::Mat(), cv::Point(-1,-1), 2 );
	}
	else
	{
		//cv::erode(mask1, mask1, element, cv::Point(1,1), 2);
		//cv::erode(mask2, mask2, element, cv::Point(1,1), 2);

		cv::erode(mask1,mask1,NULL, cv::Point(-1,-1), 2);
		cv::erode(mask2,mask2,NULL, cv::Point(-1,-1), 2);
	}

	//cvNamedWindow("Imagem da mascara1",CV_WINDOW_AUTOSIZE);
	//cvShowImage("Imagem da mascara1", mask1);

	//cvNamedWindow("Imagem da mascara2",CV_WINDOW_AUTOSIZE);
	//cvShowImage("Imagem da mascara2", mask2);

	return std::make_pair(mask1, mask2);
}