摘要
边缘检测和提取是机器视觉中必须要解决的问题,广泛应用于图像分割和模式识别等领域,由于其处理过程中经常会出现边缘不连续的情况,很多情况下并不能满足相应的工业标准。对此,根据图像处理的相关理论,结合Sobel算子,在Matlab环境中实现了对整个对象完整边缘的连续提取,并通过新算子与原有算子的边缘检测和提取的实例,最后得到了满足要求的完整边缘,验证了新算法的可行性与有效性。
Edge detecting and extracting is one primary problem in machine vision which can be applied to image segmentation and pattern recognition. And quite a lot of discrete edges would appear when applying traditional operators, which would not meet its industrial standards respectively. Based on image processing theories, this paper presented one method which can perfectly extract the whole continuous edge of one graphical object based on Sobel operator in Matlab. Finally ideal processed image with complete outline was obtained which efficiently justified the flexibility and efficiency of the processing algorithm.
出处
《机械与电子》
2011年第2期59-61,共3页
Machinery & Electronics
基金
上海市教委重点学科建设资助项目(J50503)
上海市研究生创新基金资助项目(JWCXSL0902)
