基于深度学习和强化学习的车辆定位与识别.pdf-资料库

c2367aa0-4988-4c69-89e8-ca7dfd799530.pdf-第1页.png

第1页 / 共76页

c2367aa0-4988-4c69-89e8-ca7dfd799530.pdf-第2页.png

第2页 / 共76页

c2367aa0-4988-4c69-89e8-ca7dfd799530.pdf-第3页.png

第3页 / 共76页

c2367aa0-4988-4c69-89e8-ca7dfd799530.pdf-第4页.png

第4页 / 共76页

c2367aa0-4988-4c69-89e8-ca7dfd799530.pdf-第5页.png

第5页 / 共76页

c2367aa0-4988-4c69-89e8-ca7dfd799530.pdf-第6页.png

第6页 / 共76页

c2367aa0-4988-4c69-89e8-ca7dfd799530.pdf-第7页.png

第7页 / 共76页

c2367aa0-4988-4c69-89e8-ca7dfd799530.pdf-第8页.png

第8页 / 共76页

朵击如成Ａ著 Ｃ ＳＣＩＥＮＣＥ  ＡＮＤ  ＴＥＣＨＮＯＬＯＧＹ ＯＦ ＣＨＩＮＡ ＵＮＩＶＥＲＳＩＴＹ ＯＦ ＥＬＥＣＴＲＯＮＩ硕 ± 学位论文 Ｉ ＭＡＳＴＥＲＴＨＥＳＩＳ  ．Ｉ震磅與 着 禪郷馈沪々妳 定位与识别 论文题目基于深度学习和强化学习的车辆  学科专化 学号 作者姓名 指昔教师孟瓣成副教授 ２０１３２１０７０５２８  控制科学与工程   Ｔ乐乐 Ｉ

独剑性声明 本人声明所呈交的学位论文是本人在导师指导下进行的研究工作 及取得的研究成果。据我所知，膝了文中特别加Ｗ标注和致谢的地方 夕ｈ论文中不包含其他人已经发表或撰写过的研究成果，也不包含为 获得电子科技大学或其它教育机构的学位或证书而使用过的材料。与 我一同工作的同志对本研究所做的任何贡献均已在论文中作了明确的 说明并表不谢意。 作者签名：了乐＇乐、  日期：知／（＾年处０ 论文使用授权 本学位论文作者完全了解电子科技大学有关保留、使用学位论文 的规定，有权保留并向国家有关部口或机构送交论文的复印件和磁盘， 允许论文被查阅和借阅。本人授权电子科技大学可将学位论文的全 部或部分内容编入有关数据库进行检索，可Ｗ采用影印、缩印或扫描 等复制手段保存、汇编学位论文。 （保密的学位论文在解密后应遵守此规定） 作者签名： 了乐乐  导师签名至紐坤 ；  日期：年月今２日 ＃

分类号 UDC 注 1 密级学位论文基于深度学习和强化学习的车辆定位与识别（题名和副题名）丁乐乐（作者姓名）指导教师孟继成副教授电子科技大学成都（姓名、职称、单位名称）申请学位级别硕士提交论文日期 2016.04.28 论文答辩日期学科专业控制科学与工程 2016.05.17 学位授予单位和日期电子科技大学 2016 年 06 月日答辩委员会主席评阅人注 1：注明《国际十进分类法 UDC》的类号。

VEHICLE LOCATION AND IDENTIFICATION BASED ON DEEP LEARNING AND REINFORCEMENT LEARNING A Master Thesis Submitted to University of Electronic Science and Technology of China Major: Author: Advisor: School: Control Science and Engineering Ding Lele Associate Professor Meng Jicheng School of Automation Engineering

摘要摘要车辆的定位与识别技术是智能交通中的一个关键技术，是目标车辆搜寻、车辆速度监控、车辆跟踪的必要前提。本文主要研究对于静态图片如何将深度学习与强化学习理论应用到车辆的定位与识别中，其识别的范围是判别图像中是否有车以及属于小汽车、大巴车、卡车三类中的哪一类。针对定位，使用了金字塔图像扫描时预识别出车辆并保存的位置信息，最终在识别窗口进行合并处理时定位到车辆。本文的主要内容为：首先，分析了深度学习中两个主要的网络模型：卷积神经网络以及深度信念网络。介绍了两个网络模型的结构以及原理，详细分析了两种模型的训练算法：梯度下降法和逐层贪婪算法。将两种模型用于车辆识别的仿真实验，实验结果表明深度神经网络在车辆识别上要好于传统的浅层网络，更进一步证明了卷积神经网络在本文车辆数据库上的识别性能优于深度信念网络。其次，引入强化学习思想到卷积神经网络中。将卷积神经网络与强化学习结合的模型应用到车辆识别中，分析了其理论基础以及 Q-学习算法。鉴于 Q-学习的核心在于 Agent 的动作选择策略，提出将车辆类别映射为动作，以此用基于动作选择的评估 Q 值来代替卷积神经网络中基于实际输出与期望输出的梯度后向传播训练深度神经网络。为解决卷积神经网络在训练的末尾阶段由于被错误分类的样本所占比例较小，导致对网络权值更新调整影响力较小的问题，引入基于错分样本动态调整训练集的强化学习策略。实验结果表明引入的强化学习策略可以有效的提高车辆识别的性能，并有利于网络的训练。最后，基于上述理论以及实验，设计实现了一个车辆定位与识别的实用系统。系统提供人机交互界面、深度网络的训练与测试、车辆的定位与识别等功能，并采用了多线程处理技术。测试结果表明系统能较好的识别出道路环境下小汽车、大巴车、卡车，能满足设计的需求。关键词：车辆定位与识别，深度学习，强化学习，卷积神经网络 I

ABSTRACT ABSTRACT The technology of vehicle location and identification is a one of the key technology in intelligent transportation system.It is the essential prerequisite of the target vehicle searching,the vehicle speed monitoring and the vehicle tracking.The research object of this paper is how to employ the idea of deep learning and reinforcement learning into vehicle location and identification in a static image.The target of identification is to determine whether vehicle existing and vehicle belong to which category,such as car,bus and truck.The vehicle can be located by using the position information of pyramid image technology.This paper is organized as follows: First, the two major network models are analyzed:deep belief network and convolutional neural network.The structure and principle of the two models are presented and training algorithms are analyzed in detail: gradient descent algorithm and greedy algorithm.The two models are used for the simulation experiment of vehicle identification.The simulation results indicate that deep network is better than traditional shallow network in vehicle identification and also further proof that convolution neural network performance well on the vehicle database in this article than deep belief network. Second, the idea of reinforcement learning is used to the convolutional neural network.The combination model of convolutional neural network and reinforcement learning is applied to vehicle identification.The theoretical basis of the combination model and Q-learning algorithm are analyzed.In view of the core of Q-learning lies in the action selection strategy of the Agent,we proposed the idea of vehicle categories mapping for the actions.So we can training the deep neural network by using the action-based selecting evaluation Q value instead the gradient of output error to spread in the convolutional neural network.For dealing the issue at the end of the training procedure in convolutional neural network which due to the small sample proportion of misclassification and lead to a small influence to update network weights,we proposed a reinforcement learning strategy which dynamically adjust the training set based on the misclassified samples.The simulation results indicate that the proposed reinforcement learning strategies can effectively improve the performance of the vehicle identification and is conducive to the training of deep neural network. II

ABSTRACT Finally, based on the theory and experiments designed to achieve A practical system for vehicle location and identification is designed and implemented based on the above theory and experiments.The system uses the multi thread processing technology and provides functions such as interactive interface,raining and testing of the deep network, vehicle location and identification,etc.The results indicate that the system can identify cars, buses, trucks better under the road environment.The design requirements can also be satisfied. Keywords: Vehicle location and identification, Deep Learning, Reinforcement Learning, Convolutional Neural Network III

目录目录第一章绪论................................................................................................................... 1 1.1 研究工作的背景与意义..................................................................................... 1 1.2 相关研究的国内外历史现状............................................................................. 2 1.2.1 车辆定位与识别的研究现状................................................................... 2 1.2.2 深度学习与强化学习的研究现状........................................................... 2 1.3 本文的主要贡献与创新..................................................................................... 4 1.4 本论文的结构安排............................................................................................. 5 第二章深度学习与强化学习......................................................................................... 7 2.1 深度学习............................................................................................................. 7 2.1.1 卷积神经网络模型................................................................................... 7 2.1.2 深度信念网络模型................................................................................. 10 2.2 强化学习........................................................................................................... 13 2.3 深度学习与强化学习结合模型....................................................................... 16 2.4 本章总结........................................................................................................... 18 第三章车辆识别的仿真实现....................................................................................... 20 3.1 车辆数据库制作介绍以及样本预处理............................................................ 20 3.2 深度神经网络训练的通用讨论....................................................................... 23 3.2.1 统计学习与批量学习............................................................................. 23 3.2.2 学习率..................................................................................................... 24 3.2.3 激励函数................................................................................................. 26 3.3 基于卷积神经网络的车辆识别....................................................................... 26 3.4 基于深度信念网络的车辆识别....................................................................... 31 3.5 基于卷积神经网络与强化学习的车辆识别................................................... 38 3.6 本章总结........................................................................................................... 45 第四章车辆定位与识别的系统实现........................................................................... 46 4.1 总体设计........................................................................................................... 46 4.2 车辆的定位....................................................................................................... 48 4.3 车辆的识别....................................................................................................... 50 4.4 结果与分析....................................................................................................... 53 4.5 本章总结........................................................................................................... 58 第五章总结与展望....................................................................................................... 59 IV

资料库

基于深度学习和强化学习的车辆定位与识别.pdf

相关推荐

人工智能

热门标签

最新资料