摘要
提出了单兵移动终端离线状态下基于机器视觉的装备器材识别的方法,从图像采集、图像预处理、特征提取及数据匹配3个模块展开分析。简要介绍了手持终端图像采集的相关考量。通过畸变校正、降噪、轮廓提取等,得到器材的轮廓图像,为特征提取做好准备。选取器材的几何形状特征作为识别依据,通过ORB算法得到图像的特征向量,并进行匹配,达到识别器材的目的。实验测试数据表明,该算法具有较高的识别率和良好的实时性。
In this paper,a method for identifying equipment based on Machine Vision is proposed in the off-line state of single soldier mobile terminal,which is analyzed from three modules:image acquisition,image preprocessing,feature extraction and data matching. First of all,the article briefly introduces the related considerations of handheld terminal image acquisition. Then,through distortion correction,noise reduction,contour extraction and so on,the article obtains the outline image of the equipment,and prepares for the feature extraction. By selecting the geometric features of the equipment as the recognition basis,the article uses the ORB algorithm to get the feature vectors of the images and matches then so as to achieve the purpose of identifying the equipment. The experimental data show that the algorithm has high recognition rate and good real-time performance.
作者
张建新
黄少罗
刘立家
卜昭锋
ZHANG Jian-xin;HUANG Shao-luo;LIU Li-jia;BU Zhao-feng(Shijiazhuang Campus of Army Engineering University,Shijiazhuang 050003,China;Unit 66172 of PLA,Shijiazhuang 050200,China)
出处
《火力与指挥控制》
CSCD
北大核心
2020年第2期155-159,165,共6页
Fire Control & Command Control
关键词
移动终端
机器视觉
器材识别
形状特征
ORB
mobile terminal
machine vision
equipment identification
shape feature
ORB
