Automatic Target Recognition in Complex Scene Based on Infrared Sensor
Due to different membership degrees of target solving regions in complex scenes,the target recognition accuracy is low.An auto-matic target recognition method based on infrared sensors is proposed.The target feature form in the complex scene is converted to the in-frared form,which is to the infrared sensor,and the infrared image is obtained through the light feature of the target.Based on the transfor-mation characteristics of objects in complex scenes,combined with the scale invariant feature transformation method,the feature weights of different types of objects in the scene are calculated,and the membership degrees of objects in different regions are calculated.The weight values and membership values are used as a comparison reference to establish a comparison sequence.The image centroid value and the singular value vector of all pixels in the region are calculated,a sequence is formed according to the size of the singular value vector,and compared with the preset target parameters,the pixel points that conform to the features and the expression of singular value changes are extracted,and efficient target recognition is completed.The experimental results show that the maximum value of the loss function of the proposed method is only 1.3,and the minimum value of recognition accuracy is more than 97.8％,indicating that the recognition results of the proposed method are highly consistent with the target,and the recognition results have high resolution,clear and intuitive images.
computer science and technologytarget identificationinfrared sensorscale-invariant feature transformation method
万方数据
维普
