智能工厂下基于AGV的路径识别与检测  

Path recognition and detection based on AGV in smart factories

作　　者：杨莹莹 刘翔[1] 石蕴玉[1] 黄立新 王雄飞 YANG Yingying;LIU Xiang;SHI Yunyu;HUANG Lixin;WANG Xiongfei(School of Electronic and Electrical Engineering,Shanghai University of Engineering Science,Shanghai 201620,China;Xing Jian Artificial Intelligence Technology(Shanghai)Co Ltd,Shanghai 200050,China;Nanjing Chemical Construction Co Ltd,Nanjing 211500,China)

机构地区：[1]上海工程技术大学电子电气工程学院,上海201620 [2]行謇人工智能科技(上海)有限责任公司,上海200050 [3]南京南化建设有限公司,江苏南京211500

出　　处：《传感器与微系统》2022年第9期148-150,160,共4页Transducer and Microsystem Technologies

基　　金：上海市科委自然科学基金面上资助项目(19ZR1421500);上海市科委地方能力建设项目(15590501300)。

摘　　要：为解决智能工厂中自动导引车(AGV)路径识别算法在光照不均匀、路径污染等情况下表现效果不佳的问题,提出了一种改进的KSW熵法提高路径提取效果。首先,采用LAB颜色模型对输入图像的颜色向量进行空间转换,并选取纹理及光照对路径影响最小的颜色通道作为阈值分割样本;其次,利用KSW熵法获取图像分割阈值,并结合改进的p-tile算法获取最优路径提取效果。同时,本文提出将Sobel算子的水平方向与垂直方向的边缘数据进行融合,从而降低算法的计算复杂度,并利用滑动窗口算法动态搜索边缘特征点确定AGV的行驶路径。实验结果表明:本文算法实现了路径图像在光照及污染情况下的准确提取,并且提高了路径检测的实时性与鲁棒性。In order to solve the problem in intelligent factory of path recognition algorithm for automatic guided vehicle(AGV)is interfered in the cases of uneven illumination, path pollution and so on, an improved method of KSW entropy is proposed to improve the effect of path extraction.Firstly, LAB color model is used for space transformation of color vector of the input image.Then, color channel is chosen as the sample of threshold segmentation, which has minimal impact on the path in terms of the texture and illumination.Secondly, the threshold of image segmentation is obtained by KSW entropy algorithm, which combines improved p-tile algorithm to obtain the optimal path extraction effect.Meanwhile, to reduce the computational complexity of the algorithm, fusing the horizontal and vertical edge data of Sobel operator.And, the driving path of AGV is determined by using sliding window algorithm to search edge feature points dynamically.The experimental results show that the proposed algorithm can accurately extract the path image under illumination and pollution conditions, and improve the real-time and robustness of path detection.

关 键 词：LAB颜色模型 KSW熵算法 p-tile算法 SOBEL算子 滑动窗口 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]