一种应用于高分辨率遥感图像目标检测的尺度自适应卷积神经网络
A Scale Adaptive Convolutional Neural Network for Object Detection in Remote Sensing Images
-
摘要: 为了解决高分辨率遥感图像中的目标检测问题并提高检测准确率, 本文提出了一种基于卷积神经网络的检测方法.该方法采用Faster R-CNN作为遥感图像目标检测的基础框架, 在此基础之上, 针对遥感图像中不同尺度物体的准确率差异较大这一问题, 提出了一种尺度自适应卷积设计网络.经过在遥感图像数据集上的测试, 该方法与已有的方法相比, 大大提高了高分辨率遥感图像中目标检测的准确率.Abstract: A new approach based on convolutional neural network (CNN) is proposed to improve the accuracy of multi-object in remote sensing images. Faster region convolutional neural network has been employed as the basic framework. Moreover, a scale adaptive convolutional neural network (SA-CNN) is proposed to deal with the multi-scale target detection in remote sensing images. The comparative experimental results show that the proposed SA-CNN significantly improves the accuracy of multi-object detection.