基于機(jī)器視覺(jué)的零件特征尺寸提取算法
發(fā)布時(shí)間:2018-10-30 19:55
【摘要】:目的復(fù)雜型面零件功能特征多樣,結(jié)構(gòu)尺寸呈現(xiàn)空間分布,傳統(tǒng)的手工檢測(cè)方法無(wú)法滿足檢測(cè)工作要求,為提升檢測(cè)效率,提出一種基于機(jī)器視覺(jué)的非接觸式測(cè)量方法。方法使用CCD相機(jī)采集圖像信息,對(duì)圖像進(jìn)行分析處理,獲得圓的亞像素邊緣輪廓,再通過(guò)最小二乘法進(jìn)行圓擬合求得圓的參數(shù)方程,最后利用幾何距離公式求得像素距離。通過(guò)系統(tǒng)標(biāo)定求出像素當(dāng)量,由像素當(dāng)量最終求得圓與圓之間的實(shí)際距離。結(jié)果最小二乘擬合圓亞像素邊緣檢測(cè)算法穩(wěn)定,抗噪性能較好,算法的分辨率為0.001個(gè)像素。結(jié)論該方法可正確、方便、有效地對(duì)零件進(jìn)行尺寸測(cè)量。
[Abstract]:In order to improve the detection efficiency, a non-contact measurement method based on machine vision is proposed. Methods the image information was collected by CCD camera, the image was analyzed and processed, the sub-pixel edge contour of the circle was obtained, then the circle parameter equation was obtained by the least square method, and the pixel distance was obtained by the geometric distance formula. The pixel equivalent is calculated by the system calibration, and the actual distance between the circle and the circle is obtained from the pixel equivalent. Results the edge detection algorithm of least square fitting circle subpixel is stable and robust. The resolution of the algorithm is 0.001 pixels. Conclusion this method is accurate, convenient and effective in measuring the dimensions of parts.
【作者單位】: 浙江大學(xué)寧波理工學(xué)院;寧波六和包裝有限公司;
【基金】:國(guó)家自然科學(xué)基金(51075362) 寧波市鄞州區(qū)科技局區(qū)重大產(chǎn)業(yè)技術(shù)創(chuàng)新專(zhuān)項(xiàng)(2016G002)
【分類(lèi)號(hào)】:TP391.41
[Abstract]:In order to improve the detection efficiency, a non-contact measurement method based on machine vision is proposed. Methods the image information was collected by CCD camera, the image was analyzed and processed, the sub-pixel edge contour of the circle was obtained, then the circle parameter equation was obtained by the least square method, and the pixel distance was obtained by the geometric distance formula. The pixel equivalent is calculated by the system calibration, and the actual distance between the circle and the circle is obtained from the pixel equivalent. Results the edge detection algorithm of least square fitting circle subpixel is stable and robust. The resolution of the algorithm is 0.001 pixels. Conclusion this method is accurate, convenient and effective in measuring the dimensions of parts.
【作者單位】: 浙江大學(xué)寧波理工學(xué)院;寧波六和包裝有限公司;
【基金】:國(guó)家自然科學(xué)基金(51075362) 寧波市鄞州區(qū)科技局區(qū)重大產(chǎn)業(yè)技術(shù)創(chuàng)新專(zhuān)項(xiàng)(2016G002)
【分類(lèi)號(hào)】:TP391.41
【相似文獻(xiàn)】
相關(guān)期刊論文 前10條
1 徐中宇;寧闖;周亞洲;;一種快速亞像素邊緣檢測(cè)算法[J];吉林大學(xué)學(xué)報(bào)(理學(xué)版);2014年02期
2 孫維健;;一種新的亞像素邊緣檢測(cè)算法[J];數(shù)字技術(shù)與應(yīng)用;2014年02期
3 付鵬;高曉蓉;;基于矩的亞像素邊緣檢測(cè)算法的對(duì)比研究[J];微計(jì)算機(jī)信息;2007年18期
4 張虎;達(dá)飛鵬;李勤;邢德奎;;橢圓目標(biāo)的亞像素邊緣定位方法研究[J];光學(xué)技術(shù);2010年01期
5 林欣堂;王宗義;;一種新的快速通用二維圖像亞像素邊緣檢測(cè)算法[J];應(yīng)用科技;2010年08期
6 詹林;劉秀珍;陳科澎;;亞像素邊緣定位在光纖中應(yīng)用[J];數(shù)字技術(shù)與應(yīng)用;2010年06期
7 艾澤潭;石庚辰;代俊;;微零件圖像亞像素邊緣定位算法[J];北京理工大學(xué)學(xué)報(bào);2011年03期
8 李云;夏若安;;基于曲線擬和的亞像素邊緣檢測(cè)[J];重慶科技學(xué)院學(xué)報(bào)(自然科學(xué)版);2008年06期
9 田原Z,
本文編號(hào):2301009
本文鏈接:http://sikaile.net/kejilunwen/ruanjiangongchenglunwen/2301009.html
最近更新
教材專(zhuān)著