main.cpp

/*
 *  main.cpp
 *  ExploringSfMWithOpenCV
 *
 *  Created by Roy Shilkrot on 4/27/12.
 *  Copyright 2012 MIT. All rights reserved.
 *
 */

#include <iostream>
#include <string.h>

#include "Distance.h"
#include "MultiCameraPnP.h"
#include "Visualization.h"

using namespace std;

#ifdef HAVE_OPENCV_GPU
#include <opencv2/gpu/gpu.hpp>
#endif

class VisualizerListener : public SfMUpdateListener {
public:
	void update(std::vector<cv::Point3d> pcld,
				std::vector<cv::Vec3b> pcldrgb, 
				std::vector<cv::Point3d> pcld_alternate,
				std::vector<cv::Vec3b> pcldrgb_alternate, 
				std::vector<cv::Matx34d> cameras) {
		ShowClouds(pcld, pcldrgb, pcld_alternate, pcldrgb_alternate);
		
		vector<cv::Matx34d> v = cameras;
		for(unsigned int i=0;i<v.size();i++) {
			stringstream ss; ss << "camera" << i;
			cv::Matx33f R; 
			R(0,0)=v[i](0,0); R(0,1)=v[i](0,1); R(0,2)=v[i](0,2);
			R(1,0)=v[i](1,0); R(1,1)=v[i](1,1); R(1,2)=v[i](1,2);
			R(2,0)=v[i](2,0); R(2,1)=v[i](2,1); R(2,2)=v[i](2,2);
			visualizerShowCamera(R,cv::Vec3f(v[i](0,3),v[i](1,3),v[i](2,3)),255,0,0,0.2,ss.str());
		}
	}
};

std::vector<cv::Mat> images;
std::vector<std::string> images_names;


void open_imgs_dir(char* dir_name, std::vector<cv::Mat>& images, std::vector<std::string>& images_names);

#ifndef NO_FLTK
//------------------------------ Using FLTK GUI ------------------------------
int runUI(int argc, char** argv);

int main(int argc, char** argv) { //test with real photos
	runUI(argc, argv);

	cv::destroyAllWindows();
	
//	RunVisualization(pointcloud, img_1_orig, img_2_orig, correspImg1Pt);
	
    return 0;
}

#else
//---------------------------- Using command-line ----------------------------

int main(int argc, char** argv) {
	if (argc < 2) {
		cerr << "USAGE: " << argv[0] << " <path_to_images> [use rich features (RICH/OF) = RICH] [use GPU (GPU/CPU) = GPU] [downscale factor = 1.0]" << endl;
		return 0;
	}
	
	double downscale_factor = 1.0;
	if(argc >= 5)
		downscale_factor = atof(argv[4]);

	open_imgs_dir(argv[1],images,images_names,downscale_factor);
	if(images.size() == 0) { 
		cerr << "can't get image files" << endl;
		return 1;
	}

	
	cv::Ptr<MultiCameraPnP> distance = new MultiCameraPnP(images,images_names,string(argv[1]));
	if(argc < 3)
		distance->use_rich_features = true;
	else
		distance->use_rich_features = (strcmp(argv[2], "RICH") == 0);
	
#ifdef HAVE_OPENCV_GPU
	if(argc < 4)
		distance->use_gpu = (cv::gpu::getCudaEnabledDeviceCount() > 0);
	else
		distance->use_gpu = (strcmp(argv[3], "GPU") == 0);
#else
	distance->use_gpu = false;
#endif
	
	cv::Ptr<VisualizerListener> visualizerListener = new VisualizerListener; //with ref-count
	distance->attach(visualizerListener);
	RunVisualizationThread();

	distance->RecoverDepthFromImages();

	//get the scale of the result cloud using PCA
	double scale_cameras_down = 1.0;
	{
		vector<cv::Point3d> cld = distance->getPointCloud();
		if (cld.size()==0) cld = distance->getPointCloudBeforeBA();
		cv::Mat_<double> cldm(cld.size(),3);
		for(unsigned int i=0;i<cld.size();i++) {
			cldm.row(i)(0) = cld[i].x;
			cldm.row(i)(1) = cld[i].y;
			cldm.row(i)(2) = cld[i].z;
		}
		cv::Mat_<double> mean;
		cv::PCA pca(cldm,mean,CV_PCA_DATA_AS_ROW);
		scale_cameras_down = pca.eigenvalues.at<double>(0) / 5.0;
		//if (scale_cameras_down > 1.0) {
		//	scale_cameras_down = 1.0/scale_cameras_down;
		//}
	}
	
	visualizerListener->update(distance->getPointCloud(),
							   distance->getPointCloudRGB(),
							   distance->getPointCloudBeforeBA(),
							   distance->getPointCloudRGBBeforeBA(),
							   distance->getCameras());
							   

	//ShowCloud(distance->getPointCloud(), 
	//		   distance->getPointCloudRGB(),
	//		   "baseline_only");
	//WaitForVisualizationThread();
	//return 1;
	
//	ShowClouds(distance->getPointCloud(), 
//			   distance->getPointCloudRGB(),
//			   distance->getPointCloudBeforeBA(),
//			   distance->getPointCloudRGBBeforeBA()
//			   );
	WaitForVisualizationThread();
}
#endif