基于光度立体法的黑瓜子表面三维重建研究

In this paper, the 3-D surface reconstruction of black melon seeds based on photometric stereo was studied, and which was then used to identify the wrinkled ones in bulk samples. Firstly, the system accuracy was calibrated by constructing a testing entity’s surface. Then, as there was great color difference between the central and the periphery area of the black melon seeds, the two parts’ surfaces were separately reconstructed, and the final surface was built by joining them together. Finally, experiments were done to verify the proposed method. It was shown that more than 90% of the wrinkled seeds could successfully be identified while the error rate is below 5%.