Research And Application Of Cultural Recognition Based On Edge Detection Algorithm

With the rapidly development of science and technology,the large-scale popularization of smart phones and the improvement of mobile terminal hardware performance,the exhibition of the museum has a brand-new experience.Image processing technology in the application of new way of experience has become a hot research topic,such as how to be more vivid,the image of the show users the cultural relic information,how to better image recognition technology application in the field of AR technology,when users use their mobile device to scan the cultural relics,the related information can quickly and accurately show in front of the user,and so on.In this paper,a cultural relic recognition system based on edge detection is implemented.After detecting the edge of the main cultural relic of an image,functions in Open CV are used to delineate the main contour,and the main body of the cultural relic is extracted for Harris corner point detection for feature extraction.Hausdorff distance is calculated for similarity measure to complete the classification and recognition of the cultural relic.The main research contents of this article are as follows:(1)Edge detection algorithm.Use the improved Canny edge detection operator to detect the main body of the cultural relics of the collected images,and extract the image of the main body of the cultural relics by segmentation.The improved Canny algorithm can not only extract the edge curve of the cultural relics well,but also protect the edge details of the cultural relics.(2)Feature extraction.In the feature extraction stage,Harris corner detection algorithm is used to acquire corner sample data set of the main object of the Chinese object in the collected image to form Harris corner feature.(3)Identification of cultural relics.The detected cultural relic corner point data points set is used as the classification template.When identifying the cultural relics to be detected,Harris corner point detection was also conducted to form the set of data points to be detected,and use the Hausdorff distance to measure the similarity between feature sets to complete the cultural relics classification.Using this algorithm,the overall recognition rate of seven types of cultural relics reaches 85%.(4)On the basis of the above work,a cultural relics recognition system based on edge detection is designed and implemented.Manually add photos or directly take the camera to collect the images of cultural relics,and conduct edge detection and Harris corner detection for different cultural relics.