M. Sun, S. Zhang, Y. Wei, and D. An
Machine vision, image processing, rice inspection, polar coordinate system
Due to the limitation of existing image processing algo rithms in the rectangular coordinate system, new algo rithms in the polar coordinate system are studied and ap plied to detect rice appearance quality. The regional pro cessing algorithms in the polar coordinate system, such as long-short axes detection, nonuniform sampling, and a spe cific region processing, are proposed. The maximal relative error as the result of long-short axes detection is low. The result of nonuniform sampling shows the algorithm can be effectively segmented rice images. The chalky region of rice is accurately determined by the algorithm of a specific region. All the results indicate that the proposed algorithms in the polar coordinate system are characterized by a fast detection speed, high accuracy and are rotationally invari ant.
Important Links:
Go Back
