摘要
为了提高对目标检测的效果,提出基于残差密集块的激光遥感图像中目标检测方法。首先,设计基于残差密集块的卷积神经网络,在设计ReLU激活函数并完成网络训练后,基于含噪激光遥感图像的初步特征提取结果,利用单个卷积展开卷积映射处理,抽取出潜在干净图像。然后,通过聚类处理的方式,得到激光遥感图像中车辆目标的显著图,再利用大律法,通过建立的特征比例关系的方式检测出其中的目标信息。实验结果表明,应用该方法有效滤除激光遥感图像中的噪声,并精准检测出激光遥感图像中的车辆目标。相比于3种传统方法,该方法检测结果均值误差的最小值仅为0.0156,说明该方法有效实现了设计预期。
In order to improve the effectiveness of object detection, a method for object detection in laser remotesensing images based on residual dense blocks is proposed. Firstly, design a convolutional neural network based on residualdense blocks. After designing the ReLU activation function and completing network training, based on the preliminaryfeature extraction results of noisy laser remote sensing images, use a single convolution to unfold the convolutionalmapping process and extract potentially clean images. Then, through clustering processing, the saliency map ofvehicle targets in the laser remote sensing image is obtained, and then the target information is detected using the establishedfeature proportion relationship using the general law. The experimental results show that the application ofthis method effectively filters out noise in laser remote sensing images and accurately detects vehicle targets in laser remotesensing images. Compared to the three traditional methods, the minimum value of the mean error of the detectionresults of this method is only 0. 015 6, indicating that this method effectively achieves the design expectations.
作者
李雪
刘悦
王青正
LI Xue;LIU Yue;WANG Qingzheng(College of Information Engineering,Kaifeng University,Kaifeng Henan 475000,China;Kaifeng Public Security Information Engineering Center,Kaifeng Henan 475000,China;College of Information Engineering,North China University of Water Conservancy and Electric Power,Zhengzhou 450046,China)
出处
《激光杂志》
CAS
北大核心
2024年第8期98-102,共5页
Laser Journal
基金
河南省科技攻关计划项目(No.222102210125)
河南省高等学校重点科研项目(No.23B520042)。
关键词
激光遥感图像
残差密集块
卷积神经网络
聚类算法
大律法
目标检测
去噪处理
laser remote sensing images
residual dense blocks
convolutional neural network
clustering algorithm
great law
target detection
denoising processing
