Cat eye target recognition method based on contour matching in night environment
-
摘要: 为了解决“猫眼”目标在夜晚环境下难识别的问题，提出了一种基于归一化中心矩的轮廓匹配“猫眼”目标识别方法。首先利用中值滤波对图像进行去噪，采用固定阈值分割完成了对图像的分割，使得“猫眼”目标与部分背景分离，使用Roberts边缘检测提取出了所有物体的边缘，最后采取了基于归一化中心矩的轮廓匹配算法，该算法不受平移和放缩的影响，提取出了图像中的所有圆形目标，并利用面积判别识别了真实目标，对识别出的目标绘制最小外接圆，利用圆心坐标对其定位。通过对不同光照强度下的“猫眼”图像进行实验与对比，验证了该方法的可行性，并通过目标识别评价指标验证了该方法的有效性。实验结果表明，该方法的全局准确率可达92.1%，可以在夜晚环境不同光照强度下成功地对“猫眼”目标进行识别。Abstract: To solve the problem that “cat’s eye” target is difficult to recognize at night, a contour matching algorithm based on normalized central moment is proposed. Firstly, the median filter is used to denoise the image, and the fixed threshold segmentation is used to complete the image segmentation, so that the “cat’s eye” target is separated from part of the background. Roberts edge detection is used to extract the edges of all targets. Finally, the contour matching algorithm based on the normalized central moment is adopted, which is not affected by translation and contraction. All the circular targets in the image are extracted, and the real targets are identified by area discrimination. The minimum peripheral circle is drawn for the identified targets, and the coordinates of the center of the circle are used to locate them. The feasibility of this method is verified by experiments and comparisons of “cat’s eye” images under different illumination intensities, and the effectiveness of this method is verified by target recognition evaluation index. Experimental results show that the global accuracy of this method can reach 92.1%, and it can successfully identify the “cat’s eye” target under different illumination intensity at night.
-
表 1 轮廓匹配和面积判别结果
Table 1. Contour matching and area discrimination results
target number matched-degree pixel area target number matched-degree pixel area target 0 0.010 698 target 11 0.013 18 target 2 0.016 20 target 12 0.016 20 target 4 0.002 14 target 13 0.007 26 target 6 0.002 14 target 16 0.020 14 target 8 0.010 39 target 19 0.013 18 target 9 0.002 14 target 21 0.013 18 
下载: 导出CSV
表 2 目标识别算法对比
Table 2. Comparison of target recognition algorithms
algorithm number of detected correctness number of detected errors accuracy rate/% time/ms Hough circle transformation 47 33 58.75 1156 roundness discrimination 67 13 83.75 1613 template matching 53 27 66.25 294 based on Hu moment contour matching 74 6 92.5 310 algorithm of this paper 74 6 92.5 227
下载: 导出CSV
表 3 目标定位坐标
Table 3. Target location coordinate
image location coordinate image (a) (785,982) image (b) (706,390) image (c) (1005,574) image (d) (400,233)
下载: 导出CSV
表 4 评价结果
Table 4. Evaluation result
detection
numberfalse detection
numbermissing
numberaccuracy/% false positives
rate/%false negatives
rate/%all images 129 4 7 92.1 10 7 evening environment 61 3 6 87.1 15 12 late night environment 68 1 1 97.1 5 2
下载: 导出CSV
-
[1] 殷科, 王良斯, 吴武明. 反狙击探测系统的发展现状及应对策略[J]. 四川兵工学报, 2010, 31(1):10-12Yin Ke, Wang Liangsi, Wu Wuming. Development status and countermeasures of anti-sniping detection system[J]. Journal of Sichuan Ordnance, 2010, 31(1): 10-12 [2] 石岚, 王宏. 国外反狙击手光电探测技术与装备[J]. 光电技术应用, 2010, 25(4):16-20 doi: 10.3969/j.issn.1673-1255.2010.04.005Shi Lan, Wang Hong. Foreign anti-sniper detection technology and equipment[J]. Electro-Optic Technology Application, 2010, 25(4): 16-20 doi: 10.3969/j.issn.1673-1255.2010.04.005 [3] Zhang Zhao, Song Dalin, Xu Bingshi, et al. Method of cat-eye effect target recognition based on dual-spectral imaging and deep learning[C]//Proceedings of SPIE 12343, 2nd International Conference on Laser, Optics and Optoelectronic Technology. 2022: 123432Z. [4] 同兰娟, 蒋晓瑜, 宋小杉, 等. 基于“猫眼效应”激光成像的目标探测[J]. 激光与红外, 2009, 39(9):982-985 doi: 10.3969/j.issn.1001-5078.2009.09.019Tong Lanjuan, Jiang Xiaoyu, Song Xiaoshan, et al. Target detection based on laser imaging with “cat eye effect”[J]. Laser & Infrared, 2009, 39(9): 982-985 doi: 10.3969/j.issn.1001-5078.2009.09.019 [5] 杨岳青, 李丽. 基于局部特征的猫眼效应目标识别方法[J]. 激光与红外, 2015, 45(5):580-583 doi: 10.3969/j.issn.1001-5078.2015.05.023Yang Yueqing, Li Li. Method of cat-eye effect target recognition based on local features[J]. Laser & Infrared, 2015, 45(5): 580-583 doi: 10.3969/j.issn.1001-5078.2015.05.023 [6] 李丽, 王兴宾, 张卫国. 基于纹理特征的“猫眼”效应目标识别方法[J]. 光子学报, 2014, 43(2):137-147Li Li, Wang Xingbin, Zhang Weiguo. A recognition method of “cat-eye” effect target based on texture character[J]. Acta Photonica Sinica, 2014, 43(2): 137-147 [7] 王洪玺, 计泽贤, 张兰勇. 基于卡尔曼滤波的目标识别跟踪与射击系统设计[J]. 兵器装备工程学报, 2022, 43(11):286-296 doi: 10.11809/bqzbgcxb2022.11.041Wang Hongxi, Ji Zexian, Zhang Lanyong. Design of target recognition tracking and attack system based on Kalman filter[J]. Journal of Ordnance Equipment Engineering, 2022, 43(11): 286-296 doi: 10.11809/bqzbgcxb2022.11.041 [8] 陈文龙, 张来线, 孙华燕, 等. 复杂场景下的猫眼目标快速识别方法[J]. 兵器装备工程学报, 2022, 43(7):45-51 doi: 10.11809/bqzbgcxb2022.07.008Chen Wenlong, Zhang Laixian, Sun Huayan, et al. Fast cat’s eye target recognition method in complex environment[J]. Journal of Ordnance Equipment Engineering, 2022, 43(7): 45-51 doi: 10.11809/bqzbgcxb2022.07.008 [9] 白兴斌, 张卓, 张振宇, 等. 一种基于智能瞄具的抗干扰“猫眼”目标探测方法[J]. 光电工程, 2021, 48:210115Bai Xingbin, Zhang Zhuo, Zhang Zhenyu, et al. An anti-interfering “cat-eye” target detection method based on intelligent sight[J]. Opto-Electronic Engineering, 2021, 48: 210115 [10] 王喆堃, 朱精果, 姜成昊, 等. 动态环境下“猫眼”目标快速识别算法研究[J]. 计算机仿真, 2020, 37(8):414-418 doi: 10.3969/j.issn.1006-9348.2020.08.089Wang Zhekun, Zhu Jingguo, Jiang Chenghao, et al. “Cat’s-eye” target quickly recognition algorithm research in dynamic environment[J]. Computer Simulation, 2020, 37(8): 414-418 doi: 10.3969/j.issn.1006-9348.2020.08.089 [11] 胡波, 高磊. 猫眼目标探测中数字化时间增益控制技术研究[J]. 光电技术应用, 2020, 35(4):22-25 doi: 10.3969/j.issn.1673-1255.2020.04.006Hu Bo, Gao Lei. Research on digital time gain control technology of cat-eye target detection[J]. Electro-Optic Technology Application, 2020, 35(4): 22-25 doi: 10.3969/j.issn.1673-1255.2020.04.006 [12] Kilik R. Histogram-based weighted median filtering used for noise reduction of digital elevation model data[J]. Acta Geodaetica et Geophysica, 2021, 56(4): 743-764. doi: 10.1007/s40328-021-00356-2 [13] Bath S K, Singh H, Singh G. Improve image-denoising by using weight based sparse matrix in term of MSE & PSNR[J]. International Journal of Engineering Research & Technology, 2017, 6(4): 1122-1125. [14] 万宝月. 基于OpenCV的图像分割算法研究及其在屈光度测量中的应用[D]. 西安: 西安电子科技大学, 2014Wan Baoyue. Image segmentation algorithm based on OpenCV and its application in diopter measurement[D]. Xi’an: Xidian University, 2014 [15] 何谦, 刘伯运. 红外图像边缘检测算法综述[J]. 红外技术, 2021, 43(3):199-207He Qian, Liu Boyun. Review of infrared image edge detection algorithms[J]. Infrared Technology, 2021, 43(3): 199-207 [16] Zhang Hao, Sun Qiyuan, Liu Zhenzhong. Augmented reality display of neurosurgery craniotomy lesions based on feature contour matching[J]. Cognitive Computation and Systems, 2021, 3(3): 221-228. doi: 10.1049/ccs2.12021 [17] 江波, 徐小力, 吴国新, 等. 轮廓Hu不变矩的工件图像匹配与识别[J]. 组合机床与自动化加工技术, 2020(9):104-107,111 doi: 10.13462/j.cnki.mmtamt.2020.09.023Jiang Bo, Xu Xiaoli, Wu Guoxin, et al. Workpiece recognition and matching based on Hu invariant moment of workpiece contour[J]. Modular Machine Tool & Automatic Manufacturing Technique, 2020(9): 104-107,111 doi: 10.13462/j.cnki.mmtamt.2020.09.023 [18] 杨林杰. 基于轮廓特征的目标匹配算法研究[D]. 武汉: 湖北工业大学, 2016Yang Linjie. Research on object matching algorithm based on contour feature[D]. Wuhan: Hubei University of Technology, 2016 [19] 梁龙营. 基于单相机的漆包线疵病检测系统研究[D]. 长春: 长春理工大学, 2020Liang Longying. Research on the detection system of enameled wire defects based on single camera[D]. Changchun: Changchun University of Science and Technology, 2020 -
点击查看大图
图(7) / 表(4)
计量
- 文章访问数: 497
- HTML全文浏览量: 187
- PDF下载量: 49
- 被引次数: 0
