基于机器视觉的大型零件尺寸测量技术研究

发布时间：2018-08-31 11:00

【摘要】：针对传统的接触式测量方法存在需要和零件接触以及测量时间和人为因素等局限性,研发出基于机器视觉的大型零件尺寸测量系统,从而满足产品的非接触、测量速度快、测量精度高、实时显示等测量要求。在基于机器视觉的大型零件尺寸测量系统中,首先对于采集的图像进行预处理,针对采集图像存在噪声等因素的影响,利用小波去噪完成对图像的平滑去噪;其次,选取SIFT(尺度不变特征转换)算法进行图像配准,选择基于加权平均的帽子函数加权平均融合算法无缝隙的完成图像融合,从而完成图像的拼接。然后,基于Canny检测算子提出了一种改进的单像素边缘检测方法用于提取零件图像边缘轮廓。最后,对提取的图像边缘轮廓进行几何量的实验测量并分析实验数据。基于机器视觉的大型零件尺寸测量系统不仅操作简单,价格低廉,而且测量上能够在误差的允许范围内快速测量出零件的尺寸。
[Abstract]:In view of the need of traditional contact measurement method and the limitation of part contact, measuring time and human factors, a large part size measurement system based on machine vision is developed, which can satisfy the non-contact of product and the measuring speed is fast. High accuracy, real-time display and other measurement requirements. In the measurement system of large parts size based on machine vision, the image is preprocessed first, and the image is smoothed by wavelet denoising because of the influence of noise and other factors. SIFT (Scale-Invariant feature conversion) algorithm is selected for image registration, and hat function weighted average fusion algorithm based on weighted average is chosen to complete image fusion without gaps. Then, an improved single pixel edge detection method based on Canny operator is proposed to extract the edge contour of part image. Finally, the edge contour of the extracted image is measured and the experimental data are analyzed. The measurement system based on machine vision is not only simple in operation and low in price, but also able to measure the dimensions of large parts within the allowable range of errors.
【作者单位】：沈阳理工大学;
【分类号】：TP391.41

【相似文献】