本文選題：車位檢測(cè) + 灰度直方圖��； 參考：《海南大學(xué)》2016年碩士論文

【摘要】：目前,在經(jīng)濟(jì)的推動(dòng)下,國(guó)民生活水平逐漸走向富裕,汽車的數(shù)量也在逐年增長(zhǎng),停車難的問(wèn)題日益嚴(yán)峻。如何通過(guò)現(xiàn)代科技,使停車位資源得到充分利用,已經(jīng)成為一個(gè)研究熱點(diǎn)。本文對(duì)智能停車場(chǎng)的國(guó)內(nèi)外研究現(xiàn)狀做了簡(jiǎn)要分析,并對(duì)目前比較常見的一些車位檢測(cè)方法做了對(duì)比。與非視頻車位檢測(cè)技術(shù)相比,利用視頻圖像進(jìn)行車位檢測(cè)具有明顯的優(yōu)越性。本文針對(duì)傳統(tǒng)的車位檢測(cè)算法的弊端,提出一種基于視頻圖像處理的檢測(cè)算法。這種方法整合車位圖像的灰度特征和紋理特征,將之作為判斷車位狀態(tài)的依據(jù)。特征的處理方式,是車位檢測(cè)技術(shù)的關(guān)鍵。提取各個(gè)車位圖像的灰度特征和紋理特征,將之組合成聯(lián)合特征向量。采用主成分分析法對(duì)車位的聯(lián)合特征向量進(jìn)行降維處理,并采用支持向量機(jī)對(duì)降維后的特征向量進(jìn)行訓(xùn)練。使用訓(xùn)練得到的模型對(duì)需要進(jìn)行車位檢測(cè)的視頻圖像分類,從而得到車位是否被占用的信息,并將得到的車位信息實(shí)時(shí)顯示。為了使支持向量機(jī)訓(xùn)練出更符合實(shí)際情況的模型,在訓(xùn)練樣本中加入了含有不同天氣、光照、行人、雜物及陰影的視頻圖像。經(jīng)過(guò)驗(yàn)證,該算法檢測(cè)準(zhǔn)確率為98.150%,能夠排除天氣、光照、車位上的行人、雜物、陰影等對(duì)檢測(cè)結(jié)果的影響。且該算法能同時(shí)處理多個(gè)車位,大大提高了停車場(chǎng)車位管理的效率。最后,本文利用MATLAB對(duì)此車位檢測(cè)方法進(jìn)行了模擬仿真,并在仿真的基礎(chǔ)上構(gòu)建了車位檢測(cè)實(shí)驗(yàn)系統(tǒng)。
[Abstract]:At present, driven by the economy, the national standard of living is gradually becoming rich, the number of cars is also increasing year by year, and the problem of parking difficulty is becoming more and more serious. How to make full use of parking space resources through modern science and technology has become a research hotspot. This paper makes a brief analysis of the domestic and international research status of intelligent parking, and compares some common methods of parking space detection. Compared with the non-video parking detection technology, the use of video images to detect parking space has obvious advantages. In this paper, a detection algorithm based on video image processing is proposed to overcome the disadvantages of the traditional parking space detection algorithm. This method integrates gray and texture features of parking space image as the basis for judging parking space state. The processing method of feature is the key of parking spot detection technology. The grayscale feature and texture feature of each parking space image are extracted and combined into a joint feature vector. The principal component analysis (PCA) is used to reduce the dimension of the joint eigenvector and the support vector machine (SVM) is used to train the dimensionally reduced eigenvector. The training model is used to classify the video images that need to be detected to get the information of whether the parking space is occupied or not and display the parking space information in real time. In order to train support vector machine (SVM) to fit the actual situation, video images with different weather, light, pedestrian, clutter and shadow were added to the training samples. The accuracy of the algorithm is 98.150, which can eliminate the influence of weather, illumination, pedestrians, sundries and shadows on the detection results. The algorithm can deal with multiple parking spaces at the same time, which greatly improves the efficiency of parking space management. Finally, this paper simulates the method of parking spot detection by using MATLAB, and builds an experimental system of parking space detection on the basis of simulation.
【學(xué)位授予單位】：海南大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2016
【分類號(hào)】：TP391.41

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 李磊;;基于PCA降維的SVM人臉快速檢測(cè)方法[J];青島科技大學(xué)學(xué)報(bào)(自然科學(xué)版);2010年05期

2 宋余慶;劉博;謝軍;;基于Gabor小波變換的醫(yī)學(xué)圖像紋理特征分類[J];計(jì)算機(jī)工程;2010年11期

3 肖漢光;蔡從中;;特征向量的歸一化比較性研究[J];計(jì)算機(jī)工程與應(yīng)用;2009年22期

4 蔣大林;鄧紅麗;平_g;韋燕鳳;;基于視頻圖像的多特征車位檢測(cè)算法[J];北京工業(yè)大學(xué)學(xué)報(bào);2008年02期

5 倪國(guó)強(qiáng);沈淵婷;徐大琦;;一種基于小波PCA的高光譜圖像特征提取新方法[J];北京理工大學(xué)學(xué)報(bào);2007年07期

6 張翔,田金文,肖曉玲,柳健;支持向量機(jī)及其在醫(yī)學(xué)圖像分類中的應(yīng)用[J];信號(hào)處理;2004年02期

7 楊兆升,陳曉冬;智能化停車誘導(dǎo)系統(tǒng)有效停車泊位數(shù)據(jù)的預(yù)測(cè)技術(shù)研究[J];交通運(yùn)輸系統(tǒng)工程與信息;2003年04期

8 李界家,原寶龍,朱棟華,劉建順;智能停車場(chǎng)技術(shù)及發(fā)展趨勢(shì)[J];房材與應(yīng)用;2002年04期

9 郁梅,蔣剛毅,賀賽龍;基于路面標(biāo)記的車輛檢測(cè)和計(jì)數(shù)[J];儀器儀表學(xué)報(bào);2002年04期

10 張學(xué)工;關(guān)于統(tǒng)計(jì)學(xué)習(xí)理論與支持向量機(jī)[J];自動(dòng)化學(xué)報(bào);2000年01期

相關(guān)博士學(xué)位論文 前2條

1 雷震春;支持向量機(jī)在說(shuō)話人識(shí)別中的應(yīng)用研究[D];浙江大學(xué);2006年

2 陳峻;城市停車設(shè)施規(guī)劃方法研究[D];東南大學(xué);2000年

相關(guān)碩士學(xué)位論文 前2條

1 張郝;基于小波變換的圖像去噪方法研究[D];北京交通大學(xué);2008年

2 馬銘;基于視頻的停車場(chǎng)車位狀態(tài)監(jiān)控識(shí)別系統(tǒng)[D];天津大學(xué);2007年

，

本文編號(hào)：1990510