I'm working with canny edge detector. I need to know the mean grey and the standard deviation of the resulted edges. I'm applying this in my research work; digital image analysis for cervical cancer detection.
The first step will clearly be to determine which grey values represent edges and which values don't.
Given the lower and upper limit of 'edges', you can simply have the next pass - applying your algorithm only to pixels that fit the first criterion. This will require you to count the pixels processed, but should be easily implemented.
I think you have got resultant image which should be in binary image (if you use Canny edge detection). So, just map the resultant image pixel coordinates with original image then collect all the pixel values. Calculate the mean and standard deviation using formulas which is available on net.