摘要
针对视频引伸计拉伸测试试验过程中的凸肩顶点定位难题,研究Harris角点检测及其改进的提取方法。通过理论分析与试验研究,说明改进的Harris算法较原算法在引伸计视频图像处理中,检测的角点分布更均匀,可有效地避免角点簇拥现象。为验证和改进算法的稳定性,对采集到的试件图片进行旋转和亮度变化,计算出平均角点重复率,证实改进的Harris算法能满足视频引伸计拉伸测试对动态跟踪测量的要求。
Be specific to the difficult position on sharp shrouded peak during the video extensometer about tensile test,research on the traditional Harris corner detection algorithm and an improved Harris algorithm are given. Through the theory analysis and the experimental study,the results show that the detected corners under the improved Harris algorithm are distributed more reasonable and avoid the phenomenon of clustering corners more efficiently than the original one among the video extensometer image processing system. Harris algorithm to detect the corners of specimen for verifying its stability is improved. The captured image of specimen with a tiny angle about ± 5° and vary brightness on images is rotated,which can be used to test average repetition rate. The improved Harris corner detection can satisfy the demand for the dynamic tracking measurement in the tensile testing of video extensometer.
作者
梁飞
尹秋霞
Liang Fei;Yin Qiuxia
出处
《工具技术》
2018年第3期139-141,共3页
Tool Engineering
基金
国家重大科学仪器开发专项(2013YQ220749)
关键词
几何特征
特征提取
角点检测
HARRIS算法
稳定性
动态跟踪
geometrical character
features extraction
corner detection
Harris algorithm
stabilization
dynamic tracking
