Department of Computer and Information Engineering, Inha University, South Korea.
Comput Biol Med. 2013 Jul;43(6):670-82. doi: 10.1016/j.compbiomed.2013.02.009. Epub 2013 Mar 21.
We eliminate similar frames from a wireless capsule endoscopy video of the human intestines to maximize spatial coverage and minimize the redundancy in images. We combine an intensity correction method with a method based an optical flow and features to detect and reduce near-duplicate images acquired during the repetitive backward and forward egomotions due to peristalsis. In experiments, this technique reduced duplicate image of 52.3% from images of the small intestine.
我们从人体肠道的无线胶囊内窥镜视频中消除相似的帧，以最大限度地提高空间覆盖率并最小化图像冗余。我们结合了强度校正方法和基于光流和特征的方法，以检测和减少由于蠕动导致的重复前后运动期间获取的近似重复图像。在实验中，该技术将小肠图像的重复图像减少了 52.3%。
