本文選題：圖像增強(qiáng)　切入點(diǎn)：自適應(yīng)　出處：《大連海事大學(xué)》2017年碩士論文　論文類型：學(xué)位論文

【摘要】：在計(jì)算機(jī)視覺(jué)領(lǐng)域,目標(biāo)跟蹤扮演者很重要的角色。優(yōu)秀的跟蹤算法可用于在監(jiān)控視頻中對(duì)人流車輛的數(shù)量統(tǒng)計(jì),甚至可用于導(dǎo)彈跟蹤,攔截不法入侵。為了提升在惡劣環(huán)境中對(duì)車輛目標(biāo)進(jìn)行跟蹤的性能,本文通過(guò)對(duì)雨雪、霧霾等物理模型以及在這些場(chǎng)景下所拍攝圖像的特點(diǎn)深入研究,在Stuck跟蹤算法的基礎(chǔ)之上,提出了一種惡劣場(chǎng)景下的快速自適應(yīng)跟蹤算法,并將此算法應(yīng)用到多目標(biāo)的跟蹤過(guò)程。本文的主要工作如下:在雨雪、霧霾天氣下所拍攝的圖像會(huì)出現(xiàn)降質(zhì),對(duì)比度低,模糊等現(xiàn)象,本文通過(guò)實(shí)驗(yàn)分別對(duì)比使用多尺度Retinex、暗通道先驗(yàn)去霧算法以及優(yōu)化的對(duì)比度增強(qiáng)算法進(jìn)行圖像增強(qiáng)后的效果,并分別統(tǒng)計(jì)不同場(chǎng)景及分辨率下的算法耗時(shí)。為了解決跟蹤過(guò)程中車輛目標(biāo)的尺度變化并結(jié)合原始Struck算法中生成候選樣本的方式,本文提出了基于多尺度因子數(shù)組的自適應(yīng)跟蹤算法。通過(guò)調(diào)節(jié)數(shù)組中尺度因子的數(shù)量和變化的范圍,可以生成不同數(shù)量和尺度的候選樣本,使跟蹤窗能夠自適應(yīng)車輛大小,提高跟蹤的精度。雖然改進(jìn)后的自適應(yīng)跟蹤算法能夠提高跟蹤的精度,但是對(duì)數(shù)量龐大的候選樣本進(jìn)行特征計(jì)算、分類器評(píng)估與更新等操作又降低了算法的實(shí)時(shí)性,為了將此算法應(yīng)用到多目標(biāo)跟蹤的場(chǎng)景中,因此本文基于CUDA架構(gòu)對(duì)算法進(jìn)行并行優(yōu)化,將耗時(shí)的過(guò)程放置在GPU上運(yùn)行。通過(guò)對(duì)原始算法的各個(gè)模塊的并行優(yōu)化與改進(jìn),算法的運(yùn)行速度較之前提升了三倍。針對(duì)原始算法僅適用于單目標(biāo)以及對(duì)遮擋處理的魯棒性較低的問(wèn)題,本文通過(guò)初始化多個(gè)跟蹤器并加入多線程實(shí)現(xiàn)了多個(gè)目標(biāo)的跟蹤,使得Struck算法能夠得以更廣泛的應(yīng)用。使用改進(jìn)后的算法在公共車輛圖像數(shù)據(jù)集上進(jìn)行測(cè)試,跟蹤的結(jié)果表明,改進(jìn)后的算法能夠解決跟蹤過(guò)程中車輛目標(biāo)的尺度變化以及短暫的目標(biāo)遮擋問(wèn)題。相對(duì)于原始目標(biāo)跟蹤框架,改進(jìn)后的Struck算法在跟蹤精度和速度上都有了很大提高。
[Abstract]:In the field of computer vision, target tracking plays an important role. Excellent tracking algorithms can be used to measure the number of people and vehicles in surveillance videos, and even to track missiles. Intercepting illegal intrusion. In order to improve the performance of tracking vehicle targets in harsh environments, this paper studies the physical models such as rain and snow, haze and the characteristics of images taken in these scenes. Based on the Stuck tracking algorithm, this paper proposes a fast adaptive tracking algorithm for bad scenes, and applies the algorithm to the tracking process of multiple targets. The main work of this paper is as follows: in the rain and snow, The images taken in haze weather will appear the phenomena of degradation, low contrast, blur and so on. In this paper, we compare the effect of image enhancement by using multi-scale Retinex, dark channel prior de-fogging algorithm and optimized contrast enhancement algorithm, respectively. In order to solve the scale change of vehicle target in tracking process and combine with the method of generating candidate samples in original Struck algorithm, the algorithm is used to calculate the time consuming of different scene and resolution. In this paper, an adaptive tracking algorithm based on multi-scale factor array is proposed. By adjusting the number and range of scale factors in the array, candidate samples with different numbers and scales can be generated, so that the tracking window can adapt to the vehicle size. Although the improved adaptive tracking algorithm can improve the tracking accuracy, but the feature calculation of a large number of candidate samples, classifier evaluation and update operations also reduce the real-time performance of the algorithm. In order to apply this algorithm to the scene of multi-target tracking, this paper optimizes the algorithm in parallel based on CUDA architecture and runs the time-consuming process on GPU. The running speed of the algorithm is three times faster than before. Aiming at the problem that the original algorithm is only suitable for single object and low robustness to occlusion processing, this paper initializes multiple trackers and adds multiple threads to realize the tracking of multiple targets. Struck algorithm can be more widely used. The improved algorithm is used to test the image data set of public vehicles. The tracking results show that, Compared with the original target tracking framework, the improved Struck algorithm can solve the problem of vehicle target scale change and temporary object occlusion. Compared with the original target tracking framework, the improved Struck algorithm can greatly improve the tracking accuracy and speed.
【學(xué)位授予單位】：大連海事大學(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4條

1 趙天云;鹿馨;王洪迅;李暉暉;胡秀華;;尺度自適應(yīng)結(jié)構(gòu)輸出目標(biāo)跟蹤算法[J];西北工業(yè)大學(xué)學(xué)報(bào);2016年04期

2 龔暢;王華君;李榮;徐平平;;基于狀態(tài)依賴檢測(cè)的視頻運(yùn)動(dòng)目標(biāo)軌跡跟蹤算法[J];現(xiàn)代電子技術(shù);2016年07期

3 張雷;王延杰;劉艷瀅;孫宏海;何舒文;;基于相關(guān)濾波器的視覺(jué)目標(biāo)跟蹤方法[J];光電子·激光;2015年07期

4 鐘錦;吳昊;;霧天視頻圖像增強(qiáng)與運(yùn)動(dòng)目標(biāo)跟蹤模型及方法[J];計(jì)算機(jī)仿真;2013年08期

相關(guān)博士學(xué)位論文 前1條

1 吳京輝;視頻監(jiān)控目標(biāo)的跟蹤與識(shí)別研究[D];北京理工大學(xué);2015年

相關(guān)碩士學(xué)位論文 前8條

1 左新偉;面向模糊視頻的車輛目標(biāo)跟蹤算法研究[D];大連海事大學(xué);2016年

2 陳凱;人機(jī)交互系統(tǒng)中的目標(biāo)跟蹤算法研究[D];長(zhǎng)安大學(xué);2015年

3 郭世旭;智能交通監(jiān)控系統(tǒng)中運(yùn)動(dòng)目標(biāo)跟蹤算法研究[D];河南科技大學(xué);2015年

4 鄭賢輝;單幅圖像去雨雪的算法研究[D];廈門大學(xué);2014年

5 于海彬;智能監(jiān)控系統(tǒng)的運(yùn)動(dòng)目標(biāo)檢測(cè)跟蹤算法研究與實(shí)現(xiàn)[D];哈爾濱工業(yè)大學(xué);2013年

6 井欣;霧雪天氣下視頻圖像清晰化技術(shù)的研究[D];武漢理工大學(xué);2013年

7 錢志華;多目標(biāo)跟蹤的遮擋問(wèn)題研究[D];廣東工業(yè)大學(xué);2011年

8 梁曦;基于視頻圖像的運(yùn)動(dòng)目標(biāo)檢測(cè)與跟蹤技術(shù)研究[D];西安電子科技大學(xué);2011年

，

本文編號(hào)：1612660