Demo 2D

In this demo, we consider the case of binary variables with attractive potentials. The original binary image Smile.png is degraded with noise, as showed at image Smile_noise.png. Using the DGM we restore the original image from its noised version and evaluate the results by comparing the restored image with the original one.

Input			Output
Smile.png	Smile_noise.png		Restored Image

This example copies the idea from the GraphCuts UGM Demo

#include "DGM.h"
using namespace DirectGraphicalModels;

int main(int argc, char *argv[]) 
{
    if (argc != 3) {
        print_help(argv[0]);
        return 0;
    }

    const unsigned int  nStates = 2;            // { true; false }

    // Reading parameters and images
    Mat       img       = imread(argv[1], 0);   if (img.empty())   printf("Can't open %s\n", argv[1]);
    Mat       noise     = imread(argv[2], 0);   if (noise.empty()) printf("Can't open %s\n", argv[2]);
    
    CGraphPairwise      graph(nStates);
    CGraphPairwiseExt   graphExt(graph, GRAPH_EDGES_GRID | GRAPH_EDGES_DIAG);
    CInferViterbi       decoder(graph);
    
    // no training
    vec_mat_t p(nStates);
    noise.convertTo(p[0], CV_32FC1, -1.0 / 255, 1.0);   // p_true  = 1 - noise / 255
    noise.convertTo(p[1], CV_32FC1, 1.0 / 255);         // p_false = noise / 255
    Mat nodePot;
    merge(p, nodePot);

    // ==================== Building and filling the graph ====================
    graphExt.setGraph(nodePot);
    graphExt.addDefaultEdgesModel(10000, 3);

    // =============================== Decoding ===============================
    Timer::start("Decoding... ");
    vec_byte_t optimalDecoding = decoder.decode(100);
    Timer::stop();
    
    // ====================== Evaluation / Visualization ======================
    noise = Mat(noise.size(), CV_8UC1, optimalDecoding.data()) * 255;
    medianBlur(noise, noise, 3);

    float error = 0;
    for (int y = 0; y < img.rows; y++)
        for (int x = 0; x < img.cols; x++)
            if (noise.at<byte>(y,x) != img.at<byte>(y,x)) error++;

    printf("Accuracy  = %.2f%%\n", 100 - 100 * error / (img.cols * img.rows));
    
    imshow("image", noise); 

    waitKey();

    return 0;
}