基于热释电红外传感器的运动目标识别方法  

Research on moving target recognition method based on pyroelectric infrared sensor

作　　者：陈印 李翔 张仁永 CHEN Yin;LI Xiang;ZHANG Renyong(Chongqing Institute of Engineering,Chongqing 400056,China)

机构地区：[1]重庆工程学院,重庆400056

出　　处：《激光杂志》2023年第2期181-185,共5页Laser Journal

基　　金：重庆市教委科学技术研究项目(No.KJZD-M202001901);重庆市教委科学技术研究计划青年项目(No.KJQN202101910)。

摘　　要：为精准地识别运动目标,研究基于热释电红外传感器的运动目标识别方法。构建热释电红外传感器模型,采集红外运动目标图像,通过菲涅尔透镜阵列分割和标记编码空间视觉范围,确定相应运动目标所在地点;将图像进行灰度化、阈值化、小波变换结合自适应增强算法增强处理,利用小波分析法提取图像特征,输入卷积神经网络,输出结果即运动目标识别结果。实验结果表明:所研究方法能够有效识别出运动目标,目标的检测概率近似100%,迭代训练的损失值较低,维持在0.2~0.4区间内,精度值较高,保持在0.8~1.0之间,提取的图像特征值最高可达到1,可实现高精度的目标识别。In order to identify moving targets accurately, the method of moving target identification based on pyroelectric infrared sensor is studied. Pyroelectric infrared sensor model is constructed, the infrared moving target images is collected, and the spatial visual range is divided and coded by the Fresnel lens array to determine the location of the corresponding moving target. The image is enhanced by grayscale, thresholding and wavelet transform combined with adaptive enhancement algorithm, and the image features are extracted by wavelet analysis, input into convolutional neural network, and output the result is the moving target recognition result. The experimental results show that the proposed method can effectively identify moving targets, the detection probability of targets is approximately 100%, the loss value of iterative training is low, which is maintained in the range of 0. 2 ~0. 4, and the accuracy value is high,which is maintained in the range of 0. 8~1. 0, and the extracted image feature value can reach up to 1 which can realize high-precision target recognition.

关 键 词：热释电红外 传感器 运动目标识别 菲涅尔透镜 小波分析法 自适应增强 

分 类 号：TN278[电子电信—物理电子学]