車輛牌照?qǐng)D像識(shí)別算法研究與實(shí)現(xiàn)

1.86萬(wàn)字
自己原創(chuàng)的畢業(yè)設(shè)計(jì)，今年最新的，僅在本站獨(dú)家提交，大家放心使用

摘要：近年來(lái)隨著國(guó)民經(jīng)濟(jì)的蓬勃發(fā)展，國(guó)內(nèi)高速公路、城市道路、停車場(chǎng)建設(shè)越來(lái)越多，對(duì)交通控制、安全管理的要求也日益提高。因此，汽車牌照識(shí)別技術(shù)在公共安全及交通管理中具有特別重要的實(shí)際應(yīng)用意義。本文對(duì)車牌識(shí)別系統(tǒng)中的車牌定位、字符分割和字符識(shí)別進(jìn)行了初步研究。所謂車牌定位，就是把車牌區(qū)域完整的從一幅具有復(fù)雜背景的車輛圖像中分割出來(lái)。它是進(jìn)行車牌識(shí)別的首要任務(wù)和關(guān)鍵技術(shù)，。同時(shí)，車牌定位的效率也直接影響著整個(gè)識(shí)別系統(tǒng)的效率，一個(gè)高效率的車牌識(shí)別系統(tǒng)首先必須是建立在高效的車牌定位算法的基礎(chǔ)之上。因此，研究與開(kāi)發(fā)車牌定位的算法具有十分重要的實(shí)用意義。對(duì)車牌定位，本文采用投影法對(duì)車牌進(jìn)行定位；在字符分割方面，本文使用閾值規(guī)則進(jìn)行字符分割；針對(duì)車牌圖像中數(shù)字字符識(shí)別的問(wèn)題，本文采用了基于BP神經(jīng)網(wǎng)絡(luò)的識(shí)別方法。在學(xué)習(xí)并掌握了數(shù)字圖像處理和模式識(shí)別的一些基本原理后，使用VC++6.0軟件利用以上原理針對(duì)車牌識(shí)別任務(wù)進(jìn)行編程。實(shí)現(xiàn)了對(duì)車牌的定位和車牌中數(shù)字字符的識(shí)別。

關(guān)鍵詞：車牌定位；字符分割；BP神經(jīng)網(wǎng)絡(luò)；車牌識(shí)別



Research and Realization of License Plate Recognition Algorithm
Abstract: In recent years, with the vigorous development of the national economy，there are more and more construct in the domestic expressway, urban road, and parking area. The requisition on the traffic control, safety management improves day by day. Therefore, license plate recognition technology has the particularly important practical application value in the public security and the traffic control. In the paper, a preliminary research was made on the license location, characters segment and characters recognition of the license plate recognition. A license plate location, is the complete license plate area from a vehicle in the image segmentation with complex background. It is the top priority of the license plate recognition and key technology. At the same time, the efficiency of license plate location directly affects the efficiency of the whole recognition system, must first be an efficient license plate recognition system is based on efficient license plate localization algorithm. Therefore, research and development of license plate location algorithm has very important practical significance.On the license location，the projection was used to locate the license plate; On the characters segmentation, the liminal rule was used to divide the characters; In order to solve the problem of the digital characters recognition in the plate, BP nerve network was used to recognize the digital characters. After studying and mastering some basic principles of the digital image processing and pattern recognition, the task of license plate recognition was programmed with VC++ 6.0 using above principles. The license location and the digital characters recognition in the license plate were implemented.

Key words: license plate localization;;Character segmentation;The BP neural network; License plate recognition


目 錄
第一章 緒論 1
1.1 車牌系統(tǒng)研究背景 1
1.2 車輛牌照識(shí)別系統(tǒng)原理 1
1.3 車輛牌照識(shí)別在國(guó)內(nèi)外研究現(xiàn)狀 2
1.4車輛牌照系統(tǒng)識(shí)別的主要方法 3
1.4.1 IC卡識(shí)別技術(shù) 3
1.4.2 條形碼識(shí)別技術(shù) 3
1.4.3 圖像處理與模式識(shí)別技術(shù) 3
1.4.4 人工BP網(wǎng)絡(luò)識(shí)別技術(shù) 4
1.5 本文主要工作及內(nèi)容安排 5
第二章 車輛牌照的定位方法 6
2.1車輛牌照的預(yù)處理 6
2.1.1 256色位圖灰度化 6
2.1.2 灰度圖像二值化 7
2.1.3 消除背景干擾去除噪聲 8
2.2 車輛牌照的定位方法簡(jiǎn)介 8
2.3 系統(tǒng)采用的定位方法 9
2.3.1 車輛牌照的水平定位 9
2.3.2 車輛牌照的垂直定位 10
2.3.3 定位方法的實(shí)現(xiàn) 12
2.4 實(shí)驗(yàn)結(jié)果分析 12
第三章 車牌字符分割 14
3.1 車牌預(yù)處理 14
3.1.1 去邊框處理 14
3.1.2 去噪聲處理 14
3.1.3 梯度銳化 16
3.1.4 傾斜調(diào)整 16
3.2 字符分割方法簡(jiǎn)介 17
3.3 系統(tǒng)采用的分割方法 19
3.3.1 算法介紹 19
3.3.2 算法的實(shí)現(xiàn) 20
3.4 字符分割實(shí)驗(yàn)結(jié)果 20
第四章 特征提取與字符識(shí)別 22
4.1 字符的特征提取 22
4.2 字符的識(shí)別方法簡(jiǎn)介 23
4.3 系統(tǒng)采用的識(shí)別方法 23
4.3.1 人工神經(jīng)網(wǎng)絡(luò)簡(jiǎn)介 23
4.3.2 BP神經(jīng)網(wǎng)絡(luò)識(shí)別車牌 24
4.3.3 BP神經(jīng)網(wǎng)絡(luò)識(shí)別算法實(shí)現(xiàn) 27
4.4 實(shí)驗(yàn)結(jié)果分析 28
第五章 總 結(jié) 30
致 謝 31
參考文獻(xiàn)： 32