【摘要】：針對(duì)傳統(tǒng)人工在檢測(cè)閥芯時(shí)的低效和低精度問(wèn)題,采用性能優(yōu)越的機(jī)器視覺(jué)軟件Halcon為平臺(tái),利用Canny算子邊緣檢測(cè)和雙線(xiàn)性插值方法得到閥芯圖像亞像素級(jí)邊緣,結(jié)合最小二乘算法進(jìn)行邊緣擬合,實(shí)現(xiàn)對(duì)閥芯快速、穩(wěn)定和較高精度的尺寸測(cè)量。實(shí)驗(yàn)結(jié)果表明:該檢測(cè)方式可以達(dá)到亞像素級(jí)的尺寸測(cè)量,測(cè)量精度高于2‰,滿(mǎn)足企業(yè)對(duì)測(cè)量精度要求,且測(cè)量過(guò)程比較穩(wěn)定和快速。
[Abstract]:In order to solve the problem of low efficiency and low precision in traditional manual detection of spool, the sub-pixel edge of spool image is obtained by using Canny operator edge detection and bilinear interpolation method on the platform of superior machine vision software Halcon. Combining the least square algorithm for edge fitting, the dimension measurement of the valve core is realized, which is fast, stable and accurate. The experimental results show that the measurement method can achieve sub-pixel size measurement, and the measurement accuracy is higher than 2 鈥，

本文編號(hào)：2456769