分层迭代Hough变换法检测微器件圆形轮廓

微器件装配过程中对微器件进行模式识别和定位是微器件装配系统中机器视觉的一项关键性技术。分层迭代Hough变换法先对原始图像进行分层预处理 ,然后对不同层的图像进行迭代运算 ,通过调节图像与存储数组的大小来有效地减小存储空间和计算量。实验结果验证了算法的有效性和合理性。对于实验所用的图像分层迭代 Hough变换法将存储量和计算量分别减小为标准 Hough的 1/ 316和 1/ 172。因此该算法是一种能够满足微器件装配精度要求的有效方法 ,可用于含噪图像中圆形轮廓的实时检测和定位方面

Layer-iterative Hough transform technique to detect circles in micro-assemblies

This paper proposes an efficient layer iteration search technique to detect patterns in micro-assemblies using machine vision. The technique reduced the storage and computing time for detecting circles in a noisy image. Variable size image and accumulator arrays greatly reduced the computing and storage requirements of the Hough transform. The results verified the effectiveness of this method. The required storage and computing time were reduced to 1/316 and 1/172 of the standard Hough transform using an experimental image. The layer-iteration search method is very suitable for detecting circles using the Hough transform method for micro-assemblies and can mect the accuracy requirements of the assembly operation. This method can also be applied to other pattern detection problems in real-time environments having time constraints.