【摘要】：由于電子商務(wù)網(wǎng)站的成功發(fā)展及網(wǎng)絡(luò)多媒體技術(shù)的迅速普及,在線(xiàn)購(gòu)物已經(jīng)成為一種方便、快捷、廉價(jià)且時(shí)尚的購(gòu)物方式,但隨之而來(lái)的是圖像數(shù)據(jù)呈幾何級(jí)數(shù)的增長(zhǎng),對(duì)如此超大規(guī)模的多媒體數(shù)據(jù)進(jìn)行有效管理,并提供迅速、準(zhǔn)確的檢索服務(wù)是一個(gè)極具挑戰(zhàn)性的課題。目前,電子購(gòu)物網(wǎng)站的搜索服務(wù)仍然依賴(lài)基于文本的搜索引擎,標(biāo)注并關(guān)聯(lián)商品的基本信息,對(duì)于用戶(hù)難以準(zhǔn)確地描述的樣式、花紋、造型等特有屬性缺少進(jìn)一步的標(biāo)注,將基于內(nèi)容的圖像自動(dòng)分類(lèi)引入電子商務(wù),緩解商品圖像數(shù)據(jù)庫(kù)的管理壓力和提高消費(fèi)者對(duì)商品的檢索效率,是當(dāng)前電子商務(wù)領(lǐng)域的迫切需求。 本文以在線(xiàn)購(gòu)物商品的圖像為基礎(chǔ),構(gòu)建了一個(gè)手工標(biāo)注商品特殊屬性的數(shù)據(jù)集,并以大量實(shí)驗(yàn)關(guān)注不同的圖像特征對(duì)商品圖像屬性的分類(lèi)檢測(cè)結(jié)果。主要的研究?jī)?nèi)容和貢獻(xiàn)如下： 首先,本文針對(duì)原始且粗略的在線(xiàn)商品圖像集,從購(gòu)物用戶(hù)最關(guān)注的色彩和款式兩個(gè)重要屬性出發(fā),基于商品圖像特性進(jìn)行了顏色、紋理和形狀分布的剖析,確定運(yùn)用HSV顏色空間對(duì)商品圖像提取顏色矩和顏色直方圖特征,并采用局部二值模式、梯度局部二值模式、二元梯度輪廓和方向梯度直方圖描述紋理信息和形狀信息聯(lián)合表達(dá)商品圖像的款式屬性,通過(guò)實(shí)驗(yàn)證明了這些特征具有的分類(lèi)性能。 其次,文中詳細(xì)介紹了不同底層特征對(duì)于商品顏色和款式屬性的分類(lèi)方法細(xì)節(jié),對(duì)兩個(gè)屬性層面的不同特征進(jìn)行特征級(jí)的融合,構(gòu)建復(fù)合的特征向量并通過(guò)實(shí)驗(yàn)檢驗(yàn)特征組合分類(lèi)的性能變化,實(shí)驗(yàn)結(jié)果表明,商品圖像的分類(lèi)準(zhǔn)確率得到了部分提升。 最后,雖然每種特征具備特有的分類(lèi)性能,但不同特征與分類(lèi)器決策的相關(guān)性沒(méi)有得到綜合利用,采用不同內(nèi)核的分類(lèi)算法針對(duì)特定特征會(huì)有突出的表現(xiàn),因此本文引入了多內(nèi)核學(xué)習(xí)方法改進(jìn)分類(lèi)決策,設(shè)計(jì)和運(yùn)用大量實(shí)驗(yàn)測(cè)試了顏色、紋理、形狀特征聯(lián)合描述商品圖像屬性的能力,對(duì)比了多組實(shí)驗(yàn)的結(jié)果并分析了特征在多核學(xué)習(xí)中的分類(lèi)性能。
[Abstract]:With the successful development of e-commerce websites and the rapid popularization of multimedia technology, online shopping has become a convenient, fast, cheap and fashionable way of shopping. It is a challenging task to manage multimedia data on such a large scale effectively and to provide fast and accurate retrieval services. At present, the search service of electronic shopping website still relies on the search engine based on text, marking and associating the basic information of the goods, and lack of further annotation for the unique attributes such as style, pattern, modeling and so on, which are difficult to describe accurately by the user. It is an urgent need to introduce the automatic classification of content-based images into electronic commerce to relieve the management pressure of commodity image database and to improve the retrieval efficiency of consumers in the field of electronic commerce. Based on the images of online shopping items, this paper constructs a data set of manually tagging the special attributes of commodities, and pays close attention to the classification and detection results of commodity image attributes by a large number of experiments. The main research contents and contributions are as follows: first of all, aiming at the original and rough online commodity image set, this paper starts from the two important attributes of color and style that the shopper pays most attention to, and carries on the color based on the commodity image characteristic. Based on the analysis of texture and shape distribution, HSV color space is used to extract color moments and color histogram features from commodity images, and local binary mode and gradient local binary mode are adopted. Binary gradient contour and directional gradient histogram are used to describe texture information and shape information to express the style attributes of commercial images. The classification performance of these features is proved by experiments. Secondly, this paper introduces the classification methods of different bottom features for commodity color and style attributes in detail, and combines the different features of the two attribute levels at the feature level. The experimental results show that the classification accuracy of commodity images has been partially improved. Finally, although each feature has its own classification performance, the correlation between different features and classifier decision is not comprehensively utilized. Therefore, this paper introduces a multi-kernel learning method to improve the classification decision, designs and uses a large number of experiments to test the ability of color, texture and shape features to describe the attributes of commodity images. The results of multi-group experiments are compared and the classification performance of features in multi-core learning is analyzed.
【學(xué)位授予單位】：西南交通大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2013
【分類(lèi)號(hào)】：TP391.41

【參考文獻(xiàn)】

相關(guān)期刊論文 前4條

1 孫林;吳相林;周莉;葉靖;付麗茹;;對(duì)扭轉(zhuǎn)魯棒的SIFT圖像匹配在商標(biāo)識(shí)別中的應(yīng)用[J];計(jì)算機(jī)工程與設(shè)計(jì);2011年04期

2 賈世杰;孔祥維;滿(mǎn)紅;;Automatic Product Image Classification with Multiple Support Vector Machine Classifiers[J];Journal of Shanghai Jiaotong University(Science);2011年04期

3 孫林;吳相林;羅松濤;周莉;張紅艷;;基于人體檢測(cè)的網(wǎng)絡(luò)商品圖像分類(lèi)算法[J];微計(jì)算機(jī)信息;2010年29期

4 唐玉昆;山嵐;;顏色和形狀特征的商品圖像檢索技術(shù)的應(yīng)用[J];計(jì)算機(jī)系統(tǒng)應(yīng)用;2012年09期

相關(guān)碩士學(xué)位論文 前4條

1 張非;膠囊內(nèi)窺鏡圖像特征提取和選擇算法研究[D];西南交通大學(xué);2011年

2 李雅琳;基于MPEG-7特征融合的商品圖像檢索[D];大連交通大學(xué);2011年

3 孫滔;基于顏色空間的圖像特征提取的研究[D];吉林大學(xué);2006年

4 鄧林;基于多尺度方向梯度直方圖的人體三維姿態(tài)估計(jì)[D];武漢科技大學(xué);2012年

，

本文編號(hào)：2273298