Potato's profile - activity

In the tutorial -> http://docs.opencv.org/doc/tutorials/...
I understand that the syntax for this function is as follows: perspectiveTransform(obj_corners, scene_corners, H);
Where obj_corners and scene_corners are vectors of Points and obj_corners contains 4 points of the object to be found and H is the homography matrix. My question is, what exactly does this function do?

I am trying to take an image stream and make the images appear as though they have been taken from a relatively simple camera like a CCD camera. I have added some Gaussian and salt pepper noise to give it a 'normal' look.

My next step is to warp the images or distort them. In real time this could happen due to temperature changes and lens defects of the camera. My goal is to artificially recreate this. Note that the distortion would be extremely minimal but present nonetheless. Ideally, I would like the distortion to be per image in the stream and over the sequence of images as well.

So far I have tried using perspectiveTransform( ) and warpPerspective( ), but I cannot seem to get the right parameter adjustments. Can anyone help me with this? I have also thought of using the fish-eye but could only find OpenCV functions that 'undistort' or fix fish-eye image distortion and nothing that would add a fish eye distortion.

Any thoughts or ideas on other approaches I could take to solve this problem? Thank you.

Try reading it with imread("imagename.tiff", 1) and save it using imwrite("imagename.png", image);

Would make -j $(nproc) be safer?

You can try looking into descriptor extraction and matching. OpenCV has many techniques like SURF, SIFT etc. that you can use for your task. It might be a good place to start.

I am using the xfeatures::SURF detectAndCompute() function to pull out descriptors from an image.

I see that the descriptors are stored in a Mat object. (Mat descriptors). On visually displaying the Mat I get a very strange image.

And on displaying the Mat to the console output, I get a sequence of extremely seemingly random floating point numbers. Can anyone explain to me what these actually represent? I am trying to improve my understanding of descriptors.

Thank you.

I was initially trying to implement something along these lines. Thank you for the code. It helped. Although using the line iterator was exactly what I needed.

This was exactly what I was looking for. Thank you.

Could you please provide your code?

I have detected the presence of lines in an image using the HoughLinesP() function. However, as an output, only the start and end points (coordinates) of the line are calculated by the function. I would however like to store all the points that fall on this line.

I initially thought of iterating from the starting coordinate to the end coordinate. But that gets quite complicated depending on the orientation of the line (horizontal, vertical, diagonal).

Any ideas on how I can achieve this?

You might have to disable OpenCL in your CMAKE statement. cmake ... -D WITH_OPENCL=OFF

That is what I wanted to know. What I am trying to achieve is, being able to represent the features in an image in a way that can be fed as inputs into a neural network. Do you have any ideas or suggestions for this?

On the same lines, Is there a way to reverse the computation? For example, take a set of descriptors and convert it into say a floating point value?

Thank you for the concise answer. The SO link was quite helpful. I might just have to crack open the code and take a look inside.

I am aware of the DescriptorExtractor class, but I wanted to know if there is a way to create my own key-points and my own descriptors. Say for example, I detect the co-ordinates of corners in an image, and I want to convert them to key-points and descriptors, How can this be done?

While researching, I have seen that descriptors, are calculated based on a 16x16 neighbourhood and using histograms. Can someone please explain how this can be achieved?

Thank you.

I have been using the SURF feature detector and trying to understand how it works. The usage of the SURF detector according to OpenCV 3 is --> surf->detectAndCompute(image, Mat(), keypoints, descriptors);

I want to know what exactly the Mat descriptors represent. On using imshow( ), a weird image with pixels of different gray scale intensities was shown. On printing to console, The matrix output ranged from different values like -0.65433e-05 etc.

Does anyone know what this represents?

What is your compile command?

You're absolutely right. When I said it, I meant in an ideal case. As a matter of fact I also worked on registering images of a starfield, and I had a hard time detecting corresponding points as there is only so much information that you can pull out from a starfield. I found that the pixel based registration worked well for me.