Research On Image Edge Detection Algorithm Based On Digital Morphology

With the development of machine vision and other technologies,new requirements are proposed for the accuracy and speed of image processing algorithm Among them,edge detection is a key step to quickly obtain the contour of target object and realize visual recognition,which has been widely concerned.Edge detection is of great significance in the field of machine vision and image processing.It is mainly applied to image processing and object edge extraction on a large scale.The most basic feature of an image is its edge,which contains a large amount of image information.Edge detection is the main research content of image processing and image analysis.There are many methods for edge detection.Therefore,it is impossible to extract edges effectively.As time goes by,a lot of type methods and new theories emerge during this period.In recent years,the most popular method is to adopt the method of mathematical morphology for image segmentation and edge extraction,as a result of the mathematical morphology method compared to other edge detection algorithm has the following advantages,in the process of image processing,able to efficiently filter out of the noise by using the morphological operator,as well as to retain the original information in the image.Secondly,the method of mathematical morphology can be effectively realized in the process of edge detection.It is relatively easy to realize in terms of software and hardware.The extracted edge is smooth and insensitive to noise.Based on the above theory,the following question mainly takes some gray images as examples,Uses mathematical morphology method to carry out image edge detection,and conducts experimental verification and analysis.The main contents are as follows:As the edge of an image contains most of the information in an image,the local brightness changes significantly.Therefore,in the process of image edge processing of extracting the outline of the for the determination of image edge and has very important significance,because of the traditional edge detection algorithms have some shortcomings,the most widely used with the Canny algorithm,when used alone,but when the Canny algorithm is more sensitive to noise,and extract the edge of the detail is not clear and complete,therefore to train the commonly used composite insulator,for example,the traditional Canny algorithm was improved,on the basis of the above algorithm,combined with two-dimensional Otsu,used in the insulator gray image,edge detection.By analyzing its hydrophobicity grade and evaluating its hydrophobicity index,the effectiveness and practicability of the algorithm are verified by experiments.The experiment shows that the precision of the composite insulator hydrophobicity image edge detection is improved by improving the traditional Canny algorithm.Traditional edge detection algorithm is able to handle,on the basis of image edge detection for image segmentation process do need information can be extracted effectively,but the traditional edge detection algorithm is also exist in the process of image processing is more sensitive to noise,the detected image clarity is not enough,the details is not complete,because of the multi-scale structural elements more algorithm can overcome the above shortcomings,the detected result details is complete,the image smooth,so will be used in combination,both in traditional operator,on the basis of the introduction of the multi-scale structure element algorithm,through the experiment in a couple of gray image,for example verification,It is proved that this method is feasible and can be applied to various scenarios.The mention of the algorithm is based on the deterministic theory,are on the edge of the gray image segmentation,because in real life,there are a lot of things have some uncertainty,relative to gray image,color image in the life also need to be certain segmentation,for edge extraction is important,because the cloud model theory is based on the uncertainty of the qualitative quantitative transformation model,this section is mainly based on the cloud model algorithm,using the HSV color space conversion,with color,the case of image preprocessing,concept jumped and edge segmentation,by normal cloud transformation.The image is analyzed and processed to extract the edge contour of the image,to verify the reliability of the algorithm.