External feature imported, do not find matches

[Vinmean]

Hi,

I want to use pre-extracted SIFT features and feed them into COLMAP. I am doing it, because I have additional information about these feature location that I want to use later.

When I import, I got the following log. The match doesn't work

==============================================================================

Feature import

==============================================================================

Processing file [1/72]

Features: 7741

Processing file [2/72]

.

.

.

Processing file [70/72]

Processing file [71/72]

Features: 7245

Processing file [72/72]

Elapsed time: 0.086 [minutes]

==============================================================================

Exhaustive feature matching

==============================================================================

Matching block [1/1, 1/1] in 30.173s

Elapsed time: 0.503 [minutes]

==============================================================================

Loading database

==============================================================================

Loading cameras... 36 in 0.000s

Loading matches... 0 in 0.000s

Loading images... 36 in 0.000s (connected 0)

Building correspondence graph... in 0.000s (ignored 0)

Elapsed time: 0.000 [minutes]

WARNING: No images with matches found in the database.

I saw a similar question on the github page.

https://github.com/colmap/colmap/issues/554

My problem is very similar. The post suggested that there might be some thing wrong with the feature vector. To my best, I followed the format mentioned in the tutorials. I have question though

1. Sorry, I don't understand this guideline in the tutorial "Note that by convention the upper left corner of an image has coordinate (0, 0) and the center of the upper left most pixel has coordinate (0.5, 0.5). ". Since pixels are floating point coordinates, they will have decimal values. So what is the significance of this guideline? I have used absolute (x,y) coordinates when I created my feature file. I suspect I might have gone wrong here.

I have attached one of the feature files that I created for reference. It would be great if you can spot any problem.

I used the below code to populate the feature files

void unpackOctave(const cv::KeyPoint& kpt, int& octave, int& layer,  float& scale)
{
    octave = kpt.octave & 255;
    layer = (kpt.octave >> 8) & 255;
    octave = octave < 128 ? octave : (-128 | octave);
    scale = octave >= 0 ? 1.f / (1 << octave) : (float)(1 << -octave);
}

void create_feature_file(std::vector<cv::KeyPoint>& img_keypoints,  cv::Mat& img_descriptors, std::string images_path,  std::string image_name)
{
    if (img_keypoints.size() != img_descriptors.rows)
        std::cout << "something wrong" << std::endl;

    std::ofstream outfile(images_path + "/images/" + image_name + ".txt");
    outfile << img_keypoints.size() << " " << img_descriptors.cols << std::endl;
    for (int i = 0; i < img_descriptors.rows; i++) {
        float scale;
        int octave, layer;
        unpackOctave(img_keypoints.at(i), octave, layer, scale);
        outfile << img_keypoints.at(i).pt.x << " " << img_keypoints.at(i).pt.y  << " " << scale << " " << img_keypoints.at(i).angle;
        for (int i = 0; i < img_descriptors.cols; i++) {
            float* descVector = img_descriptors.ptr<float>(i); // i - feature index
            outfile << " " << (unsigned int)*descVector;
        }
        outfile << std::endl;
    }
    outfile.close();
}

Please let me know how to come over this problem.

Thanks and Regards,

Deepak

[Johannes Schönberger]

You should write the descriptor values as integers between 0-255 and not as floating point values. See here for how to convert: https://github.com/colmap/colmap/blob/7180b62380e35fea330dde9a325d0c61f0d5111c/src/feature/utils.cc#L65

--
You received this message because you are subscribed to the Google Groups "COLMAP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to colmap+un...@googlegroups.com.
To post to this group, send email to col...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/colmap/621bbcf8-52fd-4df1-a946-2818b2dd2df9%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

<0.png.txt>

<2.png.txt>