Simulation study of Sobel operator and machine learning algorithm in hazard identification and detection in public places

Autor: Shuai Wu, Jianyi Liu, Xiuzhang Yang, Huan Xia
Rok vydání: 2021
Předmět:
Zdroj: 2021 IEEE International Conference on Advances in Electrical Engineering and Computer Applications (AEECA).
DOI: 10.1109/aeeca52519.2021.9574277
Popis: With the development of computer and image processing technology, intelligent image recognition method has been widely used in the field of public security. However, due to the influence of angle, light, noise and other factors, the accuracy of image recognition has been seriously reduced. To solve these problems, an improved Sobel operator and region selection image recognition method are proposed in this paper. Gauss smoothing and median filtering are used to reduce image noise. Sobel operator is used to extract image edge, and then the contour is binarized. Expansion and corrosion treatment expands and refines each area. Finally, region preference algorithm is used to locate the location of ID card. In this paper, Python language is used to simulate the ID image in the actual scene, which proves the effectiveness of the algorithm and strong environmental adaptability. It can effectively reduce noise, improve the accuracy of identification and location of hazards in complex public places, and effectively acquire the ability of risk prevention and control analysis. This technology can be widely used in the field of public safety, such as image recognition, crowd detection and so on. It can effectively strengthen the ability of accident prevention and control and emergency disposal. It has a certain application prospect and practical value.
Databáze: OpenAIRE