end_to_end_recognition.cpp

/*
 * textdetection.cpp
 *
 * A demo program of End-to-end Scene Text Detection and Recognition:
 * Shows the use of the Tesseract OCR API with the Extremal Region Filter algorithm described in:
 * Neumann L., Matas J.: Real-Time Scene Text Localization and Recognition, CVPR 2012
 *
 * Created on: Jul 31, 2014
 *     Author: Lluis Gomez i Bigorda <lgomez AT cvc.uab.es>
 */

#include "opencv2/text.hpp"
#include "opencv2/core/utility.hpp"
#include "opencv2/highgui.hpp"
#include "opencv2/imgproc.hpp"

#include <iostream>

using namespace std;
using namespace cv;
using namespace cv::text;

//Calculate edit distance between two words
size_t edit_distance(const string& A, const string& B);
size_t min(size_t x, size_t y, size_t z);
bool   isRepetitive(const string& s);
bool   sort_by_length(const string &a, const string &b);
//Draw ER's in an image via floodFill
void   er_draw(vector<Mat> &channels, vector<vector<ERStat> > &regions, vector<Vec2i> group, Mat& segmentation);

//Perform text detection and recognition and evaluate results using edit distance
int main(int argc, char* argv[])
{
    cout << endl << argv[0] << endl << endl;
    cout << "A demo program of End-to-end Scene Text Detection and Recognition: " << endl;
    cout << "Shows the use of the Tesseract OCR API with the Extremal Region Filter algorithm described in:" << endl;
    cout << "Neumann L., Matas J.: Real-Time Scene Text Localization and Recognition, CVPR 2012" << endl << endl;

    Mat image;
    if(argc>1)
        image  = imread(argv[1]);
    else
    {
        cout << "    Usage: " << argv[0] << " <input_image> [<gt_word1> ... <gt_wordN>]" << endl;
        return(0);
    }

    cout << "IMG_W=" << image.cols << endl;
    cout << "IMG_H=" << image.rows << endl;

    /*Text Detection*/

    // Extract channels to be processed individually
    vector<Mat> channels;

    Mat grey;
    cvtColor(image,grey,COLOR_RGB2GRAY);

    // Notice here we are only using grey channel, see textdetection.cpp for example with more channels
    channels.push_back(grey);
    channels.push_back(255-grey);

    double t_d = (double)getTickCount();
    // Create ERFilter objects with the 1st and 2nd stage default classifiers
    Ptr<ERFilter> er_filter1 = createERFilterNM1(loadClassifierNM1("trained_classifierNM1.xml"),8,0.00015f,0.13f,0.2f,true,0.1f);
    Ptr<ERFilter> er_filter2 = createERFilterNM2(loadClassifierNM2("trained_classifierNM2.xml"),0.5);

    vector<vector<ERStat> > regions(channels.size());
    // Apply the default cascade classifier to each independent channel (could be done in parallel)
    for (int c=0; c<(int)channels.size(); c++)
    {
        er_filter1->run(channels[c], regions[c]);
        er_filter2->run(channels[c], regions[c]);
    }
    cout << "TIME_REGION_DETECTION = " << ((double)getTickCount() - t_d)*1000/getTickFrequency() << endl;

    Mat out_img_decomposition= Mat::zeros(image.rows+2, image.cols+2, CV_8UC1);
    vector<Vec2i> tmp_group;
    for (int i=0; i<(int)regions.size(); i++)
    {
        for (int j=0; j<(int)regions[i].size();j++)
        {
            tmp_group.push_back(Vec2i(i,j));
        }
        Mat tmp= Mat::zeros(image.rows+2, image.cols+2, CV_8UC1);
        er_draw(channels, regions, tmp_group, tmp);
        if (i > 0)
            tmp = tmp / 2;
        out_img_decomposition = out_img_decomposition | tmp;
        tmp_group.clear();
    }

    double t_g = (double)getTickCount();
    // Detect character groups
    vector< vector<Vec2i> > nm_region_groups;
    vector<Rect> nm_boxes;
    erGrouping(image, channels, regions, nm_region_groups, nm_boxes,ERGROUPING_ORIENTATION_HORIZ);
    cout << "TIME_GROUPING = " << ((double)getTickCount() - t_g)*1000/getTickFrequency() << endl;



    /*Text Recognition (OCR)*/

    double t_r = (double)getTickCount();
    Ptr<OCRTesseract> ocr = OCRTesseract::create();
    cout << "TIME_OCR_INITIALIZATION = " << ((double)getTickCount() - t_r)*1000/getTickFrequency() << endl;
    string output;

    Mat out_img;
    Mat out_img_detection;
    Mat out_img_segmentation = Mat::zeros(image.rows+2, image.cols+2, CV_8UC1);
    image.copyTo(out_img);
    image.copyTo(out_img_detection);
    float scale_img  = 600.f/image.rows;
    float scale_font = (float)(2-scale_img)/1.4f;
    vector<string> words_detection;

    t_r = (double)getTickCount();

    for (int i=0; i<(int)nm_boxes.size(); i++)
    {

        rectangle(out_img_detection, nm_boxes[i].tl(), nm_boxes[i].br(), Scalar(0,255,255), 3);

        Mat group_img = Mat::zeros(image.rows+2, image.cols+2, CV_8UC1);
        er_draw(channels, regions, nm_region_groups[i], group_img);
        Mat group_segmentation;
        group_img.copyTo(group_segmentation);
        //image(nm_boxes[i]).copyTo(group_img);
        group_img(nm_boxes[i]).copyTo(group_img);
        copyMakeBorder(group_img,group_img,15,15,15,15,BORDER_CONSTANT,Scalar(0));

        vector<Rect>   boxes;
        vector<string> words;
        vector<float>  confidences;
        ocr->run(group_img, output, &boxes, &words, &confidences, OCR_LEVEL_WORD);

        output.erase(remove(output.begin(), output.end(), '\n'), output.end());
        //cout << "OCR output = \"" << output << "\" length = " << output.size() << endl;
        if (output.size() < 3)
            continue;

        for (int j=0; j<(int)boxes.size(); j++)
        {
            boxes[j].x += nm_boxes[i].x-15;
            boxes[j].y += nm_boxes[i].y-15;

            //cout << "  word = " << words[j] << "\t confidence = " << confidences[j] << endl;
            if ((words[j].size() < 2) || (confidences[j] < 51) ||
                    ((words[j].size()==2) && (words[j][0] == words[j][1])) ||
                    ((words[j].size()< 4) && (confidences[j] < 60)) ||
                    isRepetitive(words[j]))
                continue;
            words_detection.push_back(words[j]);
            rectangle(out_img, boxes[j].tl(), boxes[j].br(), Scalar(255,0,255),3);
            Size word_size = getTextSize(words[j], FONT_HERSHEY_SIMPLEX, (double)scale_font, (int)(3*scale_font), NULL);
            rectangle(out_img, boxes[j].tl()-Point(3,word_size.height+3), boxes[j].tl()+Point(word_size.width,0), Scalar(255,0,255),-1);
            putText(out_img, words[j], boxes[j].tl()-Point(1,1), FONT_HERSHEY_SIMPLEX, scale_font, Scalar(255,255,255),(int)(3*scale_font));
            out_img_segmentation = out_img_segmentation | group_segmentation;
        }

    }

    cout << "TIME_OCR = " << ((double)getTickCount() - t_r)*1000/getTickFrequency() << endl;


    /* Recognition evaluation with (approximate) Hungarian matching and edit distances */

    if(argc>2)
    {
        int num_gt_characters   = 0;
        vector<string> words_gt;
        for (int i=2; i<argc; i++)
        {
            string s = string(argv[i]);
            if (s.size() > 0)
            {
                words_gt.push_back(string(argv[i]));
                //cout << " GT word " << words_gt[words_gt.size()-1] << endl;
                num_gt_characters += (int)(words_gt[words_gt.size()-1].size());
            }
        }

        if (words_detection.empty())
        {
            //cout << endl << "number of characters in gt = " << num_gt_characters << endl;
            cout << "TOTAL_EDIT_DISTANCE = " << num_gt_characters << endl;
            cout << "EDIT_DISTANCE_RATIO = 1" << endl;
        }
        else
        {

            sort(words_gt.begin(),words_gt.end(),sort_by_length);

            int max_dist=0;
            vector< vector<int> > assignment_mat;
            for (int i=0; i<(int)words_gt.size(); i++)
            {
                vector<int> assignment_row(words_detection.size(),0);
                assignment_mat.push_back(assignment_row);
                for (int j=0; j<(int)words_detection.size(); j++)
                {
                    assignment_mat[i][j] = (int)(edit_distance(words_gt[i],words_detection[j]));
                    max_dist = max(max_dist,assignment_mat[i][j]);
                }
            }

            vector<int> words_detection_matched;

            int total_edit_distance = 0;
            int tp=0, fp=0, fn=0;
            for (int search_dist=0; search_dist<=max_dist; search_dist++)
            {
                for (int i=0; i<(int)assignment_mat.size(); i++)
                {
                    int min_dist_idx =  (int)distance(assignment_mat[i].begin(),
                                        min_element(assignment_mat[i].begin(),assignment_mat[i].end()));
                    if (assignment_mat[i][min_dist_idx] == search_dist)
                    {
                        //cout << " GT word \"" << words_gt[i] << "\" best match \"" << words_detection[min_dist_idx] << "\" with dist " << assignment_mat[i][min_dist_idx] << endl;
                        if(search_dist == 0)
                            tp++;
                        else { fp++; fn++; }

                        total_edit_distance += assignment_mat[i][min_dist_idx];
                        words_detection_matched.push_back(min_dist_idx);
                        words_gt.erase(words_gt.begin()+i);
                        assignment_mat.erase(assignment_mat.begin()+i);
                        for (int j=0; j<(int)assignment_mat.size(); j++)
                        {
                            assignment_mat[j][min_dist_idx]=INT_MAX;
                        }
                        i--;
                    }
                }
            }

            for (int j=0; j<(int)words_gt.size(); j++)
            {
                //cout << " GT word \"" << words_gt[j] << "\" no match found" << endl;
                fn++;
                total_edit_distance += (int)words_gt[j].size();
            }
            for (int j=0; j<(int)words_detection.size(); j++)
            {
                if (find(words_detection_matched.begin(),words_detection_matched.end(),j) == words_detection_matched.end())
                {
                    //cout << " Detection word \"" << words_detection[j] << "\" no match found" << endl;
                    fp++;
                    total_edit_distance += (int)words_detection[j].size();
                }
            }


            //cout << endl << "number of characters in gt = " << num_gt_characters << endl;
            cout << "TOTAL_EDIT_DISTANCE = " << total_edit_distance << endl;
            cout << "EDIT_DISTANCE_RATIO = " << (float)total_edit_distance / num_gt_characters << endl;
            cout << "TP = " << tp << endl;
            cout << "FP = " << fp << endl;
            cout << "FN = " << fn << endl;
        }
    }



    //resize(out_img_detection,out_img_detection,Size(image.cols*scale_img,image.rows*scale_img),0,0,INTER_LINEAR_EXACT);
    //imshow("detection", out_img_detection);
    //imwrite("detection.jpg", out_img_detection);
    //resize(out_img,out_img,Size(image.cols*scale_img,image.rows*scale_img),0,0,INTER_LINEAR_EXACT);
    namedWindow("recognition",WINDOW_NORMAL);
    imshow("recognition", out_img);
    waitKey(0);
    //imwrite("recognition.jpg", out_img);
    //imwrite("segmentation.jpg", out_img_segmentation);
    //imwrite("decomposition.jpg", out_img_decomposition);

    return 0;
}

size_t min(size_t x, size_t y, size_t z)
{
    return x < y ? min(x,z) : min(y,z);
}

size_t edit_distance(const string& A, const string& B)
{
    size_t NA = A.size();
    size_t NB = B.size();

    vector< vector<size_t> > M(NA + 1, vector<size_t>(NB + 1));

    for (size_t a = 0; a <= NA; ++a)
        M[a][0] = a;

    for (size_t b = 0; b <= NB; ++b)
        M[0][b] = b;

    for (size_t a = 1; a <= NA; ++a)
        for (size_t b = 1; b <= NB; ++b)
        {
            size_t x = M[a-1][b] + 1;
            size_t y = M[a][b-1] + 1;
            size_t z = M[a-1][b-1] + (A[a-1] == B[b-1] ? 0 : 1);
            M[a][b] = min(x,y,z);
        }

    return M[A.size()][B.size()];
}

bool isRepetitive(const string& s)
{
    int count = 0;
    for (int i=0; i<(int)s.size(); i++)
    {
        if ((s[i] == 'i') ||
                (s[i] == 'l') ||
                (s[i] == 'I'))
            count++;
    }
    if (count > ((int)s.size()+1)/2)
    {
        return true;
    }
    return false;
}


void er_draw(vector<Mat> &channels, vector<vector<ERStat> > &regions, vector<Vec2i> group, Mat& segmentation)
{
    for (int r=0; r<(int)group.size(); r++)
    {
        ERStat er = regions[group[r][0]][group[r][1]];
        if (er.parent != NULL) // deprecate the root region
        {
            int newMaskVal = 255;
            int flags = 4 + (newMaskVal << 8) + FLOODFILL_FIXED_RANGE + FLOODFILL_MASK_ONLY;
            floodFill(channels[group[r][0]],segmentation,Point(er.pixel%channels[group[r][0]].cols,er.pixel/channels[group[r][0]].cols),
                      Scalar(255),0,Scalar(er.level),Scalar(0),flags);
        }
    }
}

bool   sort_by_length(const string &a, const string &b){return (a.size()>b.size());}