Automatic unhulled rice grain crack detection by X-ray imaging

Crack detection of incoming paddy (unhulled rice grain) is an important step in the rice milling industry, as paddy grains with cracks severely affect the milling yield. The present method of crack detection by manually dehusking and examining kernels under light is laborious, time consuming, and highly subjective. The potential of x-ray imaging for paddy grain crack detection was investigated. Algorithms were developed for automatic detection of paddy cracks from the x-ray image. Hough transform was used to determine the cracks in the final segmented image. A graphical user interface (GUI) was developed for displaying the number of cracks in the given x-ray image.