基于深度学习的汽车仪表标识辨别系统设计
|
摘要
本文利用硬件在环仿真和图像识别技术,开发了汽车仪表标识自动辨别系统。系统的硬件为d SPACE仿真控制平台,模拟CAN报文实现仪表标识显示并由摄像头完成图像采集,通过双变滤波算法滤除图像噪声,调节图片像素点的对比度和亮度,利用图像帧差法对标识进行定位提取,借助深度学习Inception网络对标识信息进行辨别,通过修改全连接层结构以适应标识的分类,并推导了网络中误差求解公式。为了便于应用,设计了用户交互界面,测试结果中系统的准确率达到了86%以上,起到了很好的分类和辨别效果,辨别结果转换为CAN报文反馈给仿真机柜,从而实现了汽车仪表功能的半自动化测试。
An automatic identification system is developed using hardware-in-the-loop simulation and image recognition techniques. The hardware platform is d SPACE simulation cabinet and the automotive instrument cluster displays information through simulated CAN messages. The image captured by a camera is processed by bilateral filtering algorithm and the contrast and brightness of the image are adjusted. Icons can be detected by employing the image sequences processing technology. T he icons are then identified by the deep learning network. Inception network architecture is adopted in depth learning and full connection layer parameters are modified to fit the identified classification. The formula for solving the network is derived. MFC user interface is designed for application. The accuracy rate of the system has reached over 86%,which has achieved good classification and identification. The result is converted into CAN messages and fed back to the simulation cabinet to realize semi-automatic test of automotive instrument clusters.
An automatic identification system is developed using hardware-in-the-loop simulation and image recognition techniques. The hardware platform is d SPACE simulation cabinet and the automotive instrument cluster displays information through simulated CAN messages. The image captured by a camera is processed by bilateral filtering algorithm and the contrast and brightness of the image are adjusted. Icons can be detected by employing the image sequences processing technology. T he icons are then identified by the deep learning network. Inception network architecture is adopted in depth learning and full connection layer parameters are modified to fit the identified classification. The formula for solving the network is derived. MFC user interface is designed for application. The accuracy rate of the system has reached over 86%,which has achieved good classification and identification. The result is converted into CAN messages and fed back to the simulation cabinet to realize semi-automatic test of automotive instrument clusters.
引文
[1]李伯虎,柴旭东,张霖,等.面向新型人工智能系统的建模与仿真技术初步研究[J].系统仿真学报,2018,2(30):349-362.LI Bo-hu,CHAI Xu-dong,ZHANG Lin,et al.Preliminary study of modeling and simulation technology oriented to neo-type artifical intelligent systems[J].Journal of System Simulation,2018,2(30):349-362.
[2]DEBORAH M P.Application of an electronic instrument cluster as a node in a multiplexed vehicle electrical system[J].SAE International in united states,2001,1(2738):914-921.
[3]南江.基于图像处理的汽车组合仪表盘识别系统[D].吉林:吉林大学,2014.NAN Jiang.Car combination dashboard recognition system based on the image processing techniques[D].Jilin:Jilin university,2014.
[4]ALEGRIA F C,SERRA A C.Automatic calibration of analog and digital measuring instruments using computer vision[J].IEEE Transactions on Instrumentation&Measurement,2000,1(49):94-99.
[5]JAFFERY Z A,DUBEY A K.Real Time Visual Inspection System(RTVIS)for calibration of mechanical gauges[J].IEEE Trans on Image Processing,2011,6(32):841-846.
[6]岳晓峰,张娇.基于SIFT算法和改进最小二乘法的汽车仪表指针的识别[J].机械工程师,2014,12(13):161-163.YUE Xiao-feng,ZHANG Jiao.Automobile meter pointer recognition algorithm based on image processing technology[J].Mechanical Engineer,2014,12(13):161-163.
[7]邓书勤,巫玲,周杭,等.一种快速高精度的汽车仪表盘图像配准算法[J].西南科技大学学报,2017,4(32):71-77.DENG Shu-qin,WU Ling,ZHOU Hang,et al.A fast and high-precision image registration algorithm for automobile dashboard images[J].Journal of Southwest University of Science and Technology,2017,4(32):71-77.
[8]TOMASI C,MANDUCHI R.Bilateral filtering for gray and color images.Proceedings of the sixth international conference on computer vision[C]//Washington D C:IEEE Computer Society,1998:839-846.
[9]王玉灵.基于双边滤波的图像处理算法研究[D].西安:西安电子科技大学,2010.WANG Yu-ling.Study of algorithm in image processing based on the bilateral filter[D].Xi’an:Xidian University,2010.
[10]张超群.基于深度学习的字符识别[D].成都:电子科技大学,2016.ZHANG Chao-qun.Character recognition based on deep learning[D].Chengdu:University of Electronic Science and Technology of China,2016.
[11]徐步云,倪禾.自组织神经网络和K-means聚类算法的比较分析[J].新型工业化,2014,4(7):63-69.XU Bu-yun,NI He.Comparative analysis of SOM and K-means clustering[J].The Journal of New industrialization,2014,4(7):63-69.
[12]姚增伟,刘炜煌,王梓豪,等.基于卷积神经网络和长短时记忆神经网络的非特定人语音情感识别算法[J].新型工业化,2018,8(2):68-74.YAO Zeng-wei,LIU Wei-huang,Wang Zi-hao,et al.Speaker-indepedent speech emotion recognition algorithm based on convolutional neural and long-short memory neural network[J].The Journal of New industrialization,2018,8(2):68-74.
[13]莫宇琨,信昆仑,陈能.深度卷积神经网络在水表字符图像识别上的应用[J].供水技术,2017,11(5):54-57.MO Yu-kun,XIN Kun-lun,CHEN Neng.Application of depth convolutional neural network in watermeter digit image recognition[J].Water Technology,2017,11(5):54-57.
[14]SZEGEDY C,VANHOUCKE V,LOFFE S,et al.Rethinking the Inception Architecture for computer vision[J].Computer Science,2016:2818-2826.
[15]PHILLIP M C,TAPAS K T,KHOA N T,et al.Detection of high-grade small bowel obstruction on conventional radiography with convolutional neural networks[J].Abdominal radiology,2018,43(5):1120-1127.
[16]KIM D H,KINNON T M.Artificial intelligence in fracture detection:transfer learning from deep convolutional neural neteorks[J].Clinical radiology,2018,11(15):439-445.
[17]李嘉旋.Tensorflow技术解析与实战[M].北京:人民邮电出版社,2017:82-97.LI Jia-xuan.Tensorflow technology analysis and application[M].Beijing:The People's Posts and Telecommunications Press(Posts&Telecom Press),2017:82-97.
[18]火元莲,秦梅,宋亚丽.基于边缘特征和多帧差分法的运动目标检测算法[J].红外技术,2017,11(5):54-57.HUO Yuan-lian,QIN Mei,SONG Ya-li.Moving target detection method based on edge character and multiple frame difference,2017,11(5):54-57.
[19]张应辉,刘养硕.基于帧差法和背景差法的运动目标检测[J].计算机技术与发展,2017,2(27):25-28.ZHANG Ying-hui,LIU Yang-shuo.Moving object detection based on frame difference and background subtration[J].Computer technology and development,2017,2(27):25-28.
[20]张彦,周忠,吴威.一种用于运动物体检测的自适应更新背景模型[J].计算机辅助设计与图形学学报,2008,20(10):1316-1324.ZHANG Yan,ZHOU Zhong,WU Wei.Adaptive update background model for detecting moving objects[J].Journal of computer-aided design&computer graphics,2008,20(10):1316-1324.
[2]DEBORAH M P.Application of an electronic instrument cluster as a node in a multiplexed vehicle electrical system[J].SAE International in united states,2001,1(2738):914-921.
[3]南江.基于图像处理的汽车组合仪表盘识别系统[D].吉林:吉林大学,2014.NAN Jiang.Car combination dashboard recognition system based on the image processing techniques[D].Jilin:Jilin university,2014.
[4]ALEGRIA F C,SERRA A C.Automatic calibration of analog and digital measuring instruments using computer vision[J].IEEE Transactions on Instrumentation&Measurement,2000,1(49):94-99.
[5]JAFFERY Z A,DUBEY A K.Real Time Visual Inspection System(RTVIS)for calibration of mechanical gauges[J].IEEE Trans on Image Processing,2011,6(32):841-846.
[6]岳晓峰,张娇.基于SIFT算法和改进最小二乘法的汽车仪表指针的识别[J].机械工程师,2014,12(13):161-163.YUE Xiao-feng,ZHANG Jiao.Automobile meter pointer recognition algorithm based on image processing technology[J].Mechanical Engineer,2014,12(13):161-163.
[7]邓书勤,巫玲,周杭,等.一种快速高精度的汽车仪表盘图像配准算法[J].西南科技大学学报,2017,4(32):71-77.DENG Shu-qin,WU Ling,ZHOU Hang,et al.A fast and high-precision image registration algorithm for automobile dashboard images[J].Journal of Southwest University of Science and Technology,2017,4(32):71-77.
[8]TOMASI C,MANDUCHI R.Bilateral filtering for gray and color images.Proceedings of the sixth international conference on computer vision[C]//Washington D C:IEEE Computer Society,1998:839-846.
[9]王玉灵.基于双边滤波的图像处理算法研究[D].西安:西安电子科技大学,2010.WANG Yu-ling.Study of algorithm in image processing based on the bilateral filter[D].Xi’an:Xidian University,2010.
[10]张超群.基于深度学习的字符识别[D].成都:电子科技大学,2016.ZHANG Chao-qun.Character recognition based on deep learning[D].Chengdu:University of Electronic Science and Technology of China,2016.
[11]徐步云,倪禾.自组织神经网络和K-means聚类算法的比较分析[J].新型工业化,2014,4(7):63-69.XU Bu-yun,NI He.Comparative analysis of SOM and K-means clustering[J].The Journal of New industrialization,2014,4(7):63-69.
[12]姚增伟,刘炜煌,王梓豪,等.基于卷积神经网络和长短时记忆神经网络的非特定人语音情感识别算法[J].新型工业化,2018,8(2):68-74.YAO Zeng-wei,LIU Wei-huang,Wang Zi-hao,et al.Speaker-indepedent speech emotion recognition algorithm based on convolutional neural and long-short memory neural network[J].The Journal of New industrialization,2018,8(2):68-74.
[13]莫宇琨,信昆仑,陈能.深度卷积神经网络在水表字符图像识别上的应用[J].供水技术,2017,11(5):54-57.MO Yu-kun,XIN Kun-lun,CHEN Neng.Application of depth convolutional neural network in watermeter digit image recognition[J].Water Technology,2017,11(5):54-57.
[14]SZEGEDY C,VANHOUCKE V,LOFFE S,et al.Rethinking the Inception Architecture for computer vision[J].Computer Science,2016:2818-2826.
[15]PHILLIP M C,TAPAS K T,KHOA N T,et al.Detection of high-grade small bowel obstruction on conventional radiography with convolutional neural networks[J].Abdominal radiology,2018,43(5):1120-1127.
[16]KIM D H,KINNON T M.Artificial intelligence in fracture detection:transfer learning from deep convolutional neural neteorks[J].Clinical radiology,2018,11(15):439-445.
[17]李嘉旋.Tensorflow技术解析与实战[M].北京:人民邮电出版社,2017:82-97.LI Jia-xuan.Tensorflow technology analysis and application[M].Beijing:The People's Posts and Telecommunications Press(Posts&Telecom Press),2017:82-97.
[18]火元莲,秦梅,宋亚丽.基于边缘特征和多帧差分法的运动目标检测算法[J].红外技术,2017,11(5):54-57.HUO Yuan-lian,QIN Mei,SONG Ya-li.Moving target detection method based on edge character and multiple frame difference,2017,11(5):54-57.
[19]张应辉,刘养硕.基于帧差法和背景差法的运动目标检测[J].计算机技术与发展,2017,2(27):25-28.ZHANG Ying-hui,LIU Yang-shuo.Moving object detection based on frame difference and background subtration[J].Computer technology and development,2017,2(27):25-28.
[20]张彦,周忠,吴威.一种用于运动物体检测的自适应更新背景模型[J].计算机辅助设计与图形学学报,2008,20(10):1316-1324.ZHANG Yan,ZHOU Zhong,WU Wei.Adaptive update background model for detecting moving objects[J].Journal of computer-aided design&computer graphics,2008,20(10):1316-1324.