Abstract
Folding defects in the locust slice images are inevitable due to not yet perfect processing technology. To solve this problem, a precise interpolation method is proposed based on the OpenSURF matching algorithm and Shannon-Cosine wavelet interpolation. First, feature points are extracted from the sequence images with damaged slices and feature vectors are generated by SURF algorithm. Then the least square method is employed to build the spatial mapping model. To improve the quality of the spatial mapping model, some mismatched points are eliminated from initial feature matching points. Finally, the damaged slice image is restored by combining the spatial mapping model based on the matching point-pairs with Shannon-Cosine wavelet transform. Compared withtraditional methods, such as SAIST, Harmonic, Mixed median and OMP reconstruction, ours has the best performance with PSNR, RMSE and SSIM being 14.2689 dB, 2.0392 and 0.0403, respectively. This method is effective in repairing damaged or wrinkled areas in the locust slice image, which can help to build the 3D locust model in the future.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
