Question
*The code must be implemented with C++ and OpenCV.
Images Link
(Please do not resize the images, it is already resized as 512 x 512)
Portrait
http://postimg.org/image/kpv8kadwr/
Baloon.tif
http://postimg.org/image/tm60o84iz/
Convert your portrait into a grayscale image and decompose it with wavelet transform (you can choose your wavelet function and three levels is enough). Display the wavelet coefficients as an image. Reconstruct the edge map from the detail subbands of the first, second, and third levels (three such subbands per level.) Save the edge map into an image. Using morphology operator to remove the minor edge segments and reduce the width of the edges to one pixel size. Save the result image. Perform the same action (no need to save the images) to the attached image "Baloon.tif". Compute the energy, entropy and homogeneity features for both processed edge maps and calculated the differences. Calculate the local binary pattern of the grayscale portrait image and use morphology operator to put regions of similar texture into groups. Save the result image. Calculate three or more co-occurrence matrices of the grayscale portrait image and compute the energy, entropy, homogeneity, and correlation features for all the matrices. Save the results into an ASCII file. The program must be implemented with openCV. Put all your code in ONE program file. Submit the source code of your program and the output images and the ASCII file. Within the program, use comment lines to describe how to execute the program to demonstrate the successful completion of the above actions. Fail to include descriptive comments will receive 5 point deduction.
Explanation / Answer
I would go github for a solution.
