HAN Lu, XIE Hai-ming, GUO Shi-jun, et al. Application research on computer vision technology in surgical stapler inspection[J]. 2009, 28(3): 51-53. DOI: 10.13873/j.1000-97872009.03.023.
计算机视觉检测技术在吻合器检测中的应用研究
摘要
提出了一种利用计算视觉检测带手柄吻合器漏针新方法。通过采用正面光源投照的检测装置
综合运用图像处理和识别方法对利用CMOS相机采集的吻合器图像进行数学形态学
图像增强等方法处理
并提出模糊局部阈值分割算法实现吻合器漏针的自动识别
并根据面积特征判别漏针数。实验表明:该方法能够对多种型号颜色的吻合器进行稳定、准确的快速检测
检测正确率可达95.6%
平均每次检测时间为2.385 s
本方法实用性强
适合现代化生产的要求。
Abstract
A new method for detecting lacked needles of surgical stapler is developed.Illuminated by positive LED light
using all kinds of image processing and feature recognition
the needle is identified automatically using morphological operation
image enhancement and fuzzy local thresholds for segmentation;then the number of lacked needles can be calculated by area feature.The inspection results indicate that the rate of correctness can reach 95.6 %
and the average inspection space can reach 2.385 s/once
A Multiphase Level Set Framework for Image Segmentation Using the Mumford and Shah Model [J] . Luminita A. Vese,Tony F. Chan. International Journal of Computer Vision . 2002 (3)