JCrasher's profile - activity

I'm trying to create an object detector for detecting steering wheels using BoW + SVM approach. Unfortunately the following error is given when trying to predict:

OpenCV Error: Bad argument (The sample is not a valid vector) in cvPreparePredictData, file /build/buildd/opencv-2.4.8+dfsg1/modules/ml/src/inner_functions.cpp, line 1099
terminate called after throwing an instance of 'cv::Exception'
  what():  /build/buildd/opencv-2.4.8+dfsg1/modules/ml/src/inner_functions.cpp:1099: error: (-5) The sample is not a valid vector in function cvPreparePredictData

Aborted (core dumped)

To achieve my goal I do the following steps:

Detecting and extracting the features

Adding them to the BOWKMeansTrainer and then clustering;

Computing the BoW features and assigning labels;

Saving the dictionary, training data and labels to file;

#include <stdio.h>
#include <stdlib.h>
#include <string>

#include <omp.h>

#include <opencv2/opencv.hpp>
#include <opencv2/legacy/legacy.hpp>
#include <opencv2/nonfree/nonfree.hpp>
#include <fstream>
#include <iostream>
#include <string>

#include <dirent.h>
#include <unistd.h>
#include <sys/stat.h>
#include <sys/types.h>

using namespace cv;
using namespace std;

Mat compute_features(Mat image, Ptr<FeatureDetector>& detector, Ptr<DescriptorExtractor>& descriptors) {
    vector<KeyPoint> keypoints;
    Mat features;

        detector->detect(image, keypoints);
        KeyPointsFilter::retainBest(keypoints, 1500);
        descriptors->compute(image, keypoints, features);

        return features;
}

BOWKMeansTrainer addFeaturesToBOWKMeansTrainer(String dir, BOWKMeansTrainer& bowTrainer, Ptr<FeatureDetector>& detector, Ptr<DescriptorExtractor>& descriptors) {
        DIR *dp;
        struct dirent *dirp;
        struct stat filestat;

        dp = opendir(dir.c_str());

    Mat features;
    Mat img;

    string filepath;
    #pragma loop(hint_parallel(4))
        for (; (dirp = readdir(dp));) {
                filepath = dir + dirp->d_name;

                if (stat( filepath.c_str(), &filestat )) continue;
                if (S_ISDIR( filestat.st_mode ))         continue;

                cout << "Reading... " << filepath << endl;

                img = imread(filepath);

                features = compute_features(img, detector, descriptors);
                bowTrainer.add(features);
        }

        return bowTrainer;
}

void computeFeaturesWithBow(string dir, Ptr<FeatureDetector>& detector, Mat& trainingData, Mat& labels, BOWImgDescriptorExtractor& bowDE, int label) {
        DIR *dp;
        struct dirent *dirp;
        struct stat filestat;

        dp = opendir(dir.c_str());

        vector<KeyPoint> keypoints;
        Mat features;
        Mat img;

        string filepath;

        #pragma loop(hint_parallel(4))
        for (;(dirp = readdir(dp));) {
                filepath = dir + dirp->d_name;

                cout << "Reading: " << filepath << endl;

                if (stat( filepath.c_str(), &filestat )) continue;
                if (S_ISDIR( filestat.st_mode ))         continue;

                img = imread(filepath);

                detector->detect(img, keypoints);
                bowDE.compute(img, keypoints, features);
                KeyPointsFilter::retainBest(keypoints, 1500);

                trainingData.push_back(features);
                labels.push_back((float) label);
        }
}

void saveMat(string dir, string name, string extension, Mat& dictionary) {
        FileStorage fileStorage(dir + name + extension, FileStorage::WRITE);

        fileStorage << name << dictionary;
        fileStorage.release();
}


int main()
{
        initModule_nonfree();

        cout << "========== BOW Creator ==========" << endl;

        Ptr<FeatureDetector> detector = FeatureDetector::create("DynamicSURF");
        Ptr<DescriptorExtractor> descriptors = DescriptorExtractor::create("OpponentSURF");
        Ptr<DescriptorMatcher> matcher = DescriptorMatcher::create("FlannBased");

        cout << "initialized detector, descriptors and matcher..." << endl;

        TermCriteria tc(CV_TERMCRIT_ITER + CV_TERMCRIT_EPS, 10, 0.001);
    int dictionarySize = 1000;
    int retries = 1;
    int flags = KMEANS_PP_CENTERS;
    BOWKMeansTrainer bowTrainer(dictionarySize, tc, retries, flags);
    BOWImgDescriptorExtractor bowDE(descriptors, matcher);

    cout << "Adding Features to the bag of words..." << endl;
        addFeaturesToBOWKMeansTrainer("./dataset/training/positive_images/", bowTrainer, detector, descriptors);
        addFeaturesToBOWKMeansTrainer("./dataset/training/negative_images/", bowTrainer, detector, descriptors);
        cout << "Done.." << endl;

        cout << endl << "Clustering..." << endl;

        Mat dictionary = bowTrainer.cluster();
        saveMat("./BOW/", "dictionary", ".xml", dictionary);
        bowDE.setVocabulary(dictionary);

        Mat labels(0, 1, CV_32FC1);
        Mat trainingData(0, dictionarySize, CV_32FC1); 

        cout << endl << "Extract ...