本文選題：深度學(xué)習(xí) + D-kNN　； 參考：《鄭州大學(xué)》2017年碩士論文

【摘要】：人群密度估計(jì)與稠密人群計(jì)數(shù)是當(dāng)前計(jì)算機(jī)視覺領(lǐng)域的研究熱點(diǎn)之一,具有非常廣泛的應(yīng)用。隨著國家經(jīng)濟(jì)的持續(xù)高速發(fā)展,城鎮(zhèn)化不斷推進(jìn),城市人口規(guī)模越來越大,人群密集行為越來越多,由此帶來的恐怖事件、踩踏事件也日趨增多,如上海外灘踩踏事件、甘肅固原踩踏事件等。目前通過監(jiān)控視頻實(shí)現(xiàn)人群密度估計(jì)和準(zhǔn)確人群計(jì)數(shù)是一個(gè)至關(guān)重要的任務(wù),其結(jié)果對(duì)人群檢測(cè)、人群異常行為分析等有重要的參考作用。深度學(xué)習(xí)是一種由多個(gè)處理層組成的計(jì)算模型,它不需要人工標(biāo)注各種特征,可以通過學(xué)習(xí)獲得數(shù)據(jù)的多抽象層表示。近年來,深度學(xué)習(xí)方法的廣泛應(yīng)用顯著提高了語音識(shí)別、視覺目標(biāo)識(shí)別和檢測(cè)結(jié)果。卷積神經(jīng)網(wǎng)絡(luò)是目前深度學(xué)習(xí)中最為流行的學(xué)習(xí)算法,其主要優(yōu)勢(shì)體現(xiàn)在局部連接和權(quán)值共享,不僅降低了網(wǎng)絡(luò)模型的復(fù)雜度,減少了權(quán)值數(shù)量,而且這種網(wǎng)絡(luò)結(jié)構(gòu)對(duì)平抑、旋轉(zhuǎn)、傾斜、比例縮放等具有高度不變形。本文基于深度學(xué)習(xí)方法對(duì)復(fù)雜場(chǎng)景中人群密度估計(jì)與稠密人群計(jì)數(shù)問題進(jìn)行研究。稠密人群的特征為人群數(shù)量極大、場(chǎng)景透視、相鄰個(gè)體間存在嚴(yán)重的遮擋與阻塞,為有效降低上述特征對(duì)人群密度估計(jì)和人群計(jì)數(shù)帶來的影響,本文首先引入局部稠密概念,將圖像分塊,通過均勻化樣本、添加距離閾值、增加歐氏距離權(quán)值改進(jìn)kNN算法,并將D-kNN算法與灰度共生矩陣結(jié)合用于人群密度估計(jì)。均勻化樣本和設(shè)置閾值避免了因目標(biāo)場(chǎng)景與樣本之間的距離過大造成的誤判,保證了分類的性能;添加距離權(quán)值增強(qiáng)了特征的表示能力,降低了高維度特征對(duì)分類結(jié)果造成的影響。其次,本文借鑒卷積神經(jīng)網(wǎng)絡(luò)自動(dòng)提取特征和對(duì)場(chǎng)景扭曲的不變性,提出了一種LR-CNN稠密人群計(jì)數(shù)模型。LR-CNN模型可以從分割和壓縮過的圖像中提取到原圖像的有效信息;通過使用新的LR激活函數(shù)給卷積神經(jīng)網(wǎng)絡(luò)添加非線性因素,保留了部分負(fù)值,修正了數(shù)據(jù)分布,解決了ReLU訓(xùn)練時(shí)神經(jīng)元易死亡的問題;使用人群密度估計(jì)得出的稠密塊來訓(xùn)練LR-CNN稠密人群計(jì)數(shù)模型,降低了人群分布不均勻?qū)θ巳河?jì)數(shù)問題帶來的影響。為了驗(yàn)證和分析算法性能,本文采用當(dāng)前較為流行的ShanghaiTech和UCF_CC_50數(shù)據(jù)集。使用均絕對(duì)誤差(MAE)和均方誤差(MSE)作為評(píng)估算法性能的標(biāo)準(zhǔn),實(shí)驗(yàn)結(jié)果表明:本文設(shè)計(jì)的CNN計(jì)數(shù)模型在測(cè)試集上的MAE和MSE分別為:169.4,258.6;35.1,57.3;408.7,460.3;2.19,7.63;在稠密人群計(jì)數(shù)方面MAE和MSE較以往的方法有了明顯的降低,提高了計(jì)數(shù)的準(zhǔn)確率,對(duì)稠密人群計(jì)數(shù)因遮擋透視帶來的問題提供了有效的解決方法。通過實(shí)驗(yàn)測(cè)試及與其他方法的對(duì)比,在高人群密度場(chǎng)景下較以往的方法降低了均絕對(duì)誤差和均方誤差,提高了稠密人群計(jì)數(shù)的準(zhǔn)確率。
[Abstract]:Population density estimation and dense population counting are one of the hot topics in the field of computer vision, and they are widely used. With the sustained and rapid development of the national economy, urbanization continues to advance, the size of the urban population is getting larger and larger, and the crowd density is increasing. As a result, terrorist incidents and stampede incidents are also increasing, such as the Shanghai Bund stampede. The trampling incident in Guyuan, Gansu Province. At present, it is a very important task to estimate the population density and accurately count the population by surveillance video. The results are important reference for crowd detection and analysis of abnormal behavior of population. Depth learning is a computing model composed of multiple processing layers. It does not need to annotate all kinds of features manually, and it can be used to obtain multi-abstract layer representation of data. In recent years, the extensive application of depth learning methods has significantly improved the results of speech recognition, visual target recognition and detection. Convolution neural network is the most popular learning algorithm in depth learning at present. Its main advantages are local connection and weight sharing, which not only reduce the complexity of network model and reduce the number of weights, but also stabilize the network structure. Rotation, tilt, proportional scaling and so on has the height not to deform. In this paper, the problem of crowd density estimation and dense population counting in complex scenarios is studied based on the method of depth learning. In order to reduce the influence of the above characteristics on population density estimation and population count, the concept of local density is introduced in this paper. The image is divided into blocks and the range threshold is added to increase the Euclidean distance weight. The D-kNN algorithm is combined with the gray level co-occurrence matrix to estimate the population density. Homogenization of samples and setting of threshold value can avoid the misjudgment caused by the distance between the target scene and the sample, and ensure the performance of classification, and add the distance weight value to enhance the expression ability of the feature. The effects of high dimensional features on classification results were reduced. Secondly, this paper proposes a LR-CNN dense crowd counting model. LR-CNN model can extract the effective information of the original image from the segmented and compressed images by using the convolution neural network to extract the feature and the invariance of the scene distortion. By using the new LR activation function to add nonlinear factors to the convolutional neural network, some negative values are retained, the data distribution is corrected, and the problem of neuron death is solved during ReLU training. The LR-CNN dense population counting model is trained by using dense blocks estimated by population density estimation, which reduces the influence of uneven population distribution on population counting problem. In order to verify and analyze the performance of the algorithm, the current popular data sets of Shanghai Tech and UCF\ Mean absolute error (mae) and mean square error (MSE) are used as criteria for evaluating algorithm performance. The experimental results show that the mae and MSE of the CNN counting model designed in this paper are: 1 / 169.4258.6 / 35.1 / 35.1 / 57.3 / 408.7460.3/ 2.197.63 respectively. In dense population counting, mae and MSE are obviously lower than the previous methods, and the accuracy of counting is improved. It provides an effective solution to the problem caused by occlusion fluoroscopy in dense population counting. Through experiment and comparison with other methods, the absolute mean error and mean square error are reduced in high population density scenario, and the accuracy of dense population counting is improved.
【學(xué)位授予單位】：鄭州大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TP391.41

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 任國貞;江濤;;基于灰度共生矩陣的紋理提取方法研究[J];計(jì)算機(jī)應(yīng)用與軟件;2014年11期

2 鄭紅;李釗;李俊;;灰度共生矩陣的快速實(shí)現(xiàn)和優(yōu)化方法研究[J];儀器儀表學(xué)報(bào);2012年11期

3 焦蓬蓬;郭依正;劉麗娟;衛(wèi)星;;灰度共生矩陣紋理特征提取的Matlab實(shí)現(xiàn)[J];計(jì)算機(jī)技術(shù)與發(fā)展;2012年11期

4 吳剛;唐振民;程勇;朱峰;魏利華;;灰度共生矩陣紋理特征的運(yùn)動(dòng)目標(biāo)跟蹤方法[J];南京理工大學(xué)學(xué)報(bào)(自然科學(xué)版);2010年04期

5 高程程;惠曉威;;基于灰度共生矩陣的紋理特征提取[J];計(jì)算機(jī)系統(tǒng)應(yīng)用;2010年06期

6 苑麗紅;付麗;楊勇;苗靜;;灰度共生矩陣提取紋理特征的實(shí)驗(yàn)結(jié)果分析[J];計(jì)算機(jī)應(yīng)用;2009年04期

7 馮建輝;楊玉靜;;基于灰度共生矩陣提取紋理特征圖像的研究[J];北京測(cè)繪;2007年03期

8 王波;姚宏宇;李弼程;;一種有效的基于灰度共生矩陣的圖像檢索方法[J];武漢大學(xué)學(xué)報(bào)(信息科學(xué)版);2006年09期

9 趙海濤;董介春;張屹;;基于灰度共生矩陣的自適應(yīng)圖像邊緣檢測(cè)[J];微計(jì)算機(jī)信息;2006年17期

10 薄華;馬縛龍;焦李成;;圖像紋理的灰度共生矩陣計(jì)算問題的分析[J];電子學(xué)報(bào);2006年01期

相關(guān)碩士學(xué)位論文 前1條

1 陳興峰;基于內(nèi)容的遙感圖像數(shù)據(jù)庫檢索研究及實(shí)現(xiàn)[D];電子科技大學(xué);2008年

，

本文編號(hào)：2053939