本文關(guān)鍵詞：基于字典學(xué)習(xí)的車輛交通真實性評價模型　出處：《浙江大學(xué)》2017年碩士論文　論文類型：學(xué)位論文

  更多相關(guān)文章： 車輛交通 群組動畫 評價模型 字典學(xué)習(xí)

【摘要】：近些年,車輛交通群組動畫在各行各業(yè)得以廣泛應(yīng)用。關(guān)于交通模擬的研究也越來越引人注目,包括基于宏觀特征和微觀模型在內(nèi)的許多新穎的模擬方法被提了出來。在某些實際應(yīng)用中,如城市交通管理,對動畫的真實性要求非常之高。然而,關(guān)于交通模擬的真實性評價的工作卻鮮有人研究。本文提出了一個完整的車輛交通群組動畫的真實性評價模型。本模型參考了交通模擬中紋理合成的方法,將車流數(shù)據(jù)統(tǒng)一進行圖像信號化處理。首先,從大量真實車流數(shù)據(jù)中提取一些低維特征信號。然后,通過字典學(xué)習(xí)算法從真實信號中提出車流結(jié)構(gòu)性特征。最后,這些特征與測試信號進行比較,通過測試信號的重構(gòu)殘差計算出一個評價值,并以此來實現(xiàn)對車輛交通動畫真實性的評價。本文為車流數(shù)據(jù)的模擬提供了一套完善客觀的評價系統(tǒng)和反饋糾正機制,以促進這一研究方向的發(fā)展。并為車輛交通群組動畫的應(yīng)用如城市交通的治理起到積極的作用。此外,本文使用字典學(xué)習(xí)算法來評價車流數(shù)據(jù)的思想,也為整個群組動畫方向的研究開辟了一條嶄新的思路。
[Abstract]:In recent years, traffic group animation has been widely used in all walks of life. The research on traffic simulation is more and more noticeable, including many novel features and micro macro simulation method, based on the model is proposed. In some practical applications, such as city traffic management, real demand for animation is very high however. True, evaluation about traffic simulation work but few research. This paper proposes a complete evaluation of real vehicle traffic group animation model. The model reference texture synthesis method in traffic simulation, the traffic data unified signal processing image. Firstly, some low dimensional feature extraction from the signal a large number of real traffic data. Then, proposed traffic structural features from the real signal through dictionary learning. Finally, the characteristics and the test signals are compared through An evaluation value calculation of residual test signal reconstruction, in order to achieve the evaluation of vehicle traffic animation authenticity. This paper provides a complete and objective evaluation system for traffic flow data and feedback mechanism, to promote the development of the study and application. For group animation such as vehicle traffic city traffic management play a positive role. In addition, this paper use a dictionary to evaluate traffic data the idea of learning algorithm, but also opens up a new way to study the whole group animation direction.

【學(xué)位授予單位】：浙江大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：TP391.41

【參考文獻】

相關(guān)期刊論文 前2條

1 王世明;徐建閩;羅強;李日涵;;面向高速公路的車輛換道安全預(yù)警模型[J];華南理工大學(xué)學(xué)報(自然科學(xué)版);2014年12期

2 袁超;何東;于光輝;;關(guān)于振動采集設(shè)備頻率分辨率的分析[J];東方汽輪機;2013年01期

，

本文編號：1388692