tobix10's profile - activity

Get foreground mask without updating model I want to write an algorithm that detects camera tampering e.g blocking the v

Get foreground mask without updating model I want to write an algorithm that detects camera tampering e.g blocking the v

Ok, I see that there is a problem with class index. Person is a first class in a file, but 1 is subtracted from index in

Ok, I see that there is a problem with class index. Person is a first class in a file, but 1 is subtracted from it in th

Nice thanks, but it seems that this network doesn't output classes. Is this format [batchId, classId, confidence, left,

I use the model that is linked. I think that input is not a problem here e.g for such image it doesn't work either. Yolo

Do you have the same results for original yolov3-tiny from https://pjreddie.com/darknet/yolo/ ? I am experimenting with

I use the model that is linked. I think that input is not a problem here e.g for such image it doesn't work either.

Faster RCNN no results I am trying to run Faster-RCNN Inception v2 model in OpenCV 3.4.6 (c++) using object_detection.cp

That is the name of the file, because I tested yolo first. Those are coco class names. It doesn't change anything, those

Faster RCNN no results I am trying to run Faster-RCNN Inception v2 model in OpenCV 3.4.6 (c++) using object_detection.cp

Faster RCNN no results I am trying to run Faster-RCNN Inception v2 model in OpenCV 3.4.6 (c++) using object_detection.cp

Faster RCNN no results I am trying to run Faster-RCNN Inception v2 model (https://github.com/opencv/opencv/wiki/TensorFl

Camera view is blocked, uniform image Hi, how would you approach a problem of detecting if camera view is blocked? Simp

@supra56 Maybe it is, but on ARM I have a few FPS (it is paired with some morph op and find contours). At the beginning

Maybe it is, but on ARM I have a few FPS (it is paired with some morph op and find contours). At the beginning of pipeli

Find significant differences in image Hi, I'd like to make background segmentation as a first step in my processing pipe

Find significant differences in image Hi, I'd like to make background segmentation as a first step in my processing pipe

Find significant differences in image Hi, I'd like to make background segmentation as a first step in my processing pipe

Find significant differences in image Hi, I'd like to make background segmentation as a first step in my processing pipe

Yeah, but this is a profiler. I still have to write some code. I am not an expert in image processing and OpenCV. I woul

Performance analysis I would like to prepare a test which will give me an exact answer if particular CPU (ARM) is capabl

HOG groupRectangles I was searching for some information about finalThreshold parameter in detectMutliScale and found th

Hello, I'd like to know if my code is good. I want to get Fourier descriptors from contour(shape). Correct me if I am wrong: Fourier descriptors = DFT output?

I'd also want transform, scale and rotation invariance. To get this I reject F[0], divide all F[i]/F[1] and use only magnitudes |F[i]|, correct?

Here is the code(java);

Imgproc.findContours(image, contours, hierarchy, Imgproc.RETR_EXTERNAL, Imgproc.CHAIN_APPROX_NONE); 
...
Point[] arr = contours.get(maxIdx).toArray();
Mat pointsX = new Mat(new Size(1,128), CvType.CV_32FC1);
Mat pointsY = new Mat(new Size(1,128), CvType.CV_32FC1);    
int delta = arr.length / 128;             // 128 points from contour
for(int i = 0, j = 0; i < arr.length && j < 128; i += delta, ++j) {
   pointsX.put(j, 0, arr[i].x); 
   pointsY.put(j, 0, arr[i].y); 
}   

List<Mat> planes = new ArrayList<Mat>();
planes.add(pointsX);
planes.add(pointsY);
Mat complexI = new Mat();
Core.merge(planes, complexI);                   
Core.dft(complexI, complexI);               

// scale invariant Fi = Fi / |F1|
double Re = complexI.get(1,0)[0];
double Im = complexI.get(1,0)[1];
double magF1 = Math.sqrt(Re*Re + Im*Im);
for(int i = 2; i < complexI.rows(); ++i) {
   double[] newVal = new double[2];
   newVal[0] = complexI.get(i, 0)[0] / magF1;
   newVal[1] = complexI.get(i, 0)[1] / magF1;
   complexI.put(i, 0, newVal);
}
// rotation invariant |Fi|
Core.split(complexI, planes);            // planes[0] = Re(DFT(I), planes[1] = Im(DFT(I))
Mat dftMag = new Mat();
Core.magnitude(planes.get(0), planes.get(1), dftMag);

I don't know why but after I did inverse dft(just after forward dft) I got big values such (30000,4000).

Mat idft = new Mat(new Size(1,128), CvType.CV_32FC2);
Core.dft(complexI, idft, Core.DFT_INVERSE, 0);

Is this good approach? Did I miss something here?

Ok. Post edited.

I have object's bounding rect and contour. I want to use a contour as a mask to enlarge boundaries of this object to the size of bounding rect. In other words if object with green boundary pixels fitted in rectangle is on black background I want to make black pixels green.

How to do that?

This is image after thresholding and applying findContours. I want to use this contour as a mask to extract leaf from original image and expand its boundaries to fill rectangle. I need to perform texture calculations and don't want to have black pixels around.

Hello, I'm going to implement GLCM to retrieve texture features of leaves. I've already read about algorithm and it is clear to me, but I don't know:

1) How to choose the best offset of neighbours(distance and direction)?

In one text I've read that too large glcm matrix may give worse results if image has less gray levels(many zeroes in matrix). Author gives advice to rescale data to 4 bit.

2) Is it worth to do that? Will image not lost many information?

2a) Can I do rescaling in opencv with some built-in function?

Despite of above, I intended to take gray scale image and calculate GLCM on entire image, but there is also a concept of window not clear for me, so:

3) How the algoritm will lool like with using window of particular size?

Please take into account tha performance is important to me, because this will be used in android app.

4) Is there anything else important about GLCM on what should I pay attention?

Thanks in advance.