期刊
MICROPROCESSORS AND MICROSYSTEMS
卷 80, 期 -, 页码 -出版社
ELSEVIER
DOI: 10.1016/j.micpro.2020.103554
关键词
Ship target image recognition; Preprocessing; Feature extraction; Neural network (NN) classification; Infrared image
类别
资金
- Social Science Fund Project in Jiangxi Province, learning Commitment Study Based on Big Data in MOOC [17JY16]
- Science and Technology Fund Project in Jiangxi Education Department, Parllel Studying and Implementation of Face Recognition Algorithms on Video Big Data [GJJ161670]
The article introduces a classification method for infrared intrusion target detection based on neural networks, which successfully combines static and dynamic analysis, is applied to ship image analysis, and a extended neural network is designed to fuse feature images.
Infrared target ship detection is an essential application in the field of military defense and intelligent early warning. Considering the characteristics of intrusion targets and the difficulty of inspection, classification based on Neural Networks (NN) has been proposed for infrared intrusion target detection. This NN calculation joins static objective example investigation with dynamic, diverse connection discovery to remove infrared components at various levels. Among these, infrared boat picture examination can be successfully used to distinguish the back element designs in the objective library. The movement outline contrast strategy, as a camouflage, provides the integrity of the target area. To improve, it can detect the motion area of the image, to combine the benefits of these two methods, an extended convolutional neural network has been designed to fuse and extend the feature images obtained by these two methods. Network expansion modules enhance filter targets and provide background suppression for infrared images. Experimental results contrasted with customary infrared objective discovery techniques. The proposed strategy can recognize infrared interruption targets all the more precisely and smother clamor all the more viably out of sight based on FPGA Xilinx software.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据