Detection of surgical instruments based on Gaussian kernel

In minimally invasive laparoscopic surgery, it is of practical significance to quickly locate the location and category information surgical instrument. It can remind medical personnel irreversible injury caused patients due leaving instruments after operation. this paper, Gaussian kernel introduced into each ground truth, which conducive making full use label allocate positive negative samples improve accuracy classification. Then, we introduce SIoU Loss Harmonic function a total loss. The former uses relative coordinates make network converge more quickly, latter solves problem asynchronous optimization two branches classification regression. Our experiment proves that strategy based on sample allocation very effective pubic data set m2cai16-tool-locations, displaying our method possesses conspicuous regression than other work.