基于深度学习和双目相机的手术辅助目标检测与定位

In response to the need for automatic and intelligent detection and positioning of surgically assisted targets during surgery, this paper designs a surgically assisted target detection and positioning system based on YOLOv3 algorithm.This article crops the picture according to the output of the YOLOv3 algorithm, and then uses the center detection algorithm to detect the center of the surgically assisted target. Based on the detection results, the three-dimensional coordinates of the surgical auxiliary target are calculated using the parallax principle. At last, this paper completes the experimental part of the surgery-assisted target detection and positioning system to verify the effectiveness of the system.