[1]孙维亚,陈科仲,李忠新,等.基于 LBP 特征的多视点目标分辨技术研究[J].计算机技术与发展,2021,31(09):43-47.[doi:10. 3969 / j. issn. 1673-629X. 2021. 09. 008]
　SUN Wei-ya,CHEN Ke-zhong,LI Zhong-xin,et al.Research on Multi-view Target Resolution Based on LBP Characteristics[J].,2021,31(09):43-47.[doi:10. 3969 / j. issn. 1673-629X. 2021. 09. 008]
点击复制

基于 LBP 特征的多视点目标分辨技术研究()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
31
期数:
2021年09期
页码:
43-47
栏目:
图形与图像
出版日期:
2021-09-10

文章信息/Info

Title:
Research on Multi-view Target Resolution Based on LBP Characteristics
文章编号:
1673-629X(2021)09-0043-05
作者:
孙维亚1 ; 陈科仲2 ; 李忠新1 ; 吴志林1
1. 南京理工大学 机械工程学院,江苏 南京 210094;
2. 陆装装驻重庆地区军事代表局,重庆 400060
Author(s):
SUN Wei-ya1 ; CHEN Ke-zhong2 ; LI Zhong-xin1 ; WU Zhi-lin1
1. School of Mechanical Engineering,Nanjing University of Science and Technology,Nanjing 210094,China;?
2. Army Military Representative Bureau in Chongqing,Chongqing 400060,China
关键词:
光电感知系统; 目标分辨; 多视点; Haar; 局部二值模式
Keywords:
photoelectric sensing system; objective distinguishing; multiple viewpoints; Haar; LBP
分类号:
TP391
DOI:
10. 3969 / j. issn. 1673-629X. 2021. 09. 008
摘要:
在未来作战战场中,不同视点的应用相对单一视点具有更好的灵活性,对战争走向的获取、分析更为全面。 文中提出一种针对未来作战战场,不同视点下单兵进行目标分辨的方法,分析了该场景下的主要问题和难点,设计了采用光电感知系统采集视点中的目标,利用 Haar-like 算法提取视点中目标的 ROI 区域坐标,将此提取结果作为目标特征算子的计算范围并利用 LBP( local binary pattern,局部二值模式) 特征算法进行纹理计算,最后根据目标一致性判断依据进行目标分辨的试验。 试验结果表明:利用 Haar-like 算法和 LBP 算法能够有效实现目标区域坐标的提取和多视点的目标分辨,并针对 Haar-like 算法和 LBP 算法组合检测的识别精度和该方法的判断有效性分别进行测试,测试结果表明该组合算法的识别精度明显提升,且该方法的正确分辨率达到 86. 7% 。
Abstract:
In the future battlefield, the application of different viewpoints is more flexible than that of a single viewpoint, and the acquisition and analysis of the direction of the war are more comprehensive. We propose a method? of target resolution for individual soldier under different viewpoints in the future battlefield,analyze the main problems and difficulties in this scenario,and design the photoelectric sensing system to collect a viewpoint? ?of the target. The Haar-like algorithm is used to extract the ROI coordinates of the object in the viewpoint,? and the extracted results are used as the calculation range of the target feature operator and LBP (local binary pattern) feature algorithm is used to calculate the texture. Finally,the experiment of target resolution? ? ?is carried out according to the judgment basis of target consistency. The experiment shows that using Haar-like algorithm and LBP algorithm can effectively achieve the target area coordinate multiple viewpoints and? the extraction of target resolution. Moreover,the recognition accuracy of the combined detection of Haar-like algorithm and LBP algorithm and the judgment effectiveness of the proposed method are tested respectively. The test shows that the recognition accuracy of the combined algorithm is significantly improved,and the correct resolution of the method reaches 86. 7% .
更新日期/Last Update: 2021-09-10