Infrared Panoramic Image Mosaic And Target Detection Technology

The development of technology has brought about an increase in wealth and living standards,while also further increasing security requirements.Compared to traditional security equipment,infrared detection technology detects target objects by sensing their thermal signals,effectively detecting disguised targets,possessing good anti-interference ability,and being able to work all day.It has broad application prospects in the field of security.Due to the fixed detection perspective of infrared detectors,a single infrared detector cannot meet the requirements of large and wide field of view in the field of security in practical applications.Therefore,in order to improve the practicability of infrared detection technology in the security field,this paper designs an infrared panoramic image mosaic and target detection method to realize the effective monitoring of regional security by a single infrared detector.In this paper,the infrared panoramic image mosaic technology is theoretically analyzed and studied,and an image mosaic method based on improved SIFT feature extraction algorithm is designed.In the stage of image registration,ROI algorithm is used to frame select specific areas of the image,and SIFT algorithm is used to extract only the feature data in the frame selected area,reducing the overall calculation amount and improving the detection efficiency of the algorithm.The KNN classification algorithm is used to match the extracted feature points,and the efficient operation of the algorithm is realized by constructing a k-d tree.Then RANSAC algorithm is used to delete the mismatched point pairs and complete the infrared image registration.In the image fusion phase,the weighted average fusion algorithm of gradual in and gradual out is adopted,and the smooth fusion of the registered image is achieved by setting the pixel weight of the fusion part,and finally the mosaic of the infrared panoramic image is completed.This article conducts research and analysis on infrared target detection technology,and designs a target detection method based on gamma transform mixed Gaussian background modeling.Gamma transform is used to enhance the contrast of infrared image,and the mixed Gaussian background modeling method is used to establish the image background model and extract the foreground target.The accuracy of target extraction is determined by calculating the image gray distribution.Finally,the foreground target in the infrared image is selected according to the target extraction effect frame to complete the infrared target detection.In this paper,several groups of infrared images collected in different regions are used to verify the above theory.The results show that the average running speed of the algorithm in this paper has increased by 52.78%,and the average registration rate has increased by 7.49%.Compared with the infrared panoramic images obtained by the traditional algorithm,the subjective effect and objective evaluation have been significantly improved.In terms of infrared target detection,the algorithm in this paper has an average detection accuracy of 83.30% in multiple scenes,which has certain practical value.