Bright edges detector

edge = cv.BrightEdges(img)
edge = cv.BrightEdges(img, 'OptionName',optionValue, ...)

Input

• img input color image.

Output

• edge output edge image.

Options

• Contrast default 1
• ShortRange default 3
• LongRange default 9

The function implements a new way of detecting edges used is low resolution image object recognition in real projects (e.g. 50cm per pixel). It corrects surfaces for objects partially under a lighting shadow, and reveal low visibility edges. The result is used to feed further object detection processes particularly in context where small edges or details are important for object discrimination.

The function provides an implementation of an equalized absolute difference of blurs, and an optional further treatment to contrast edges based on finding local minimum along at least two directions. The local minimum detection contrast is a parameter, using 10 as default (10 on 255 maximum). The kernel size for the gaussian and the average blur can be passed as parameters too.