Image processing with opencv in c++ and hierarchical cluster

I'm looking for someone with good experience in image processing and machine learning who can help me to remove the background on a image, make lines pixels segmentation (For example we take one image with the size 100*100 which mean 100 lines for rows and 100 lines for columns. I want to extract the pixels value of each line and store them into a csv file ). One we get the pixels of each line into a csv file create an hierarchical cluster to find the similar lines.