《電子技術(shù)應(yīng)用》
您所在的位置:首頁(yè) > 其他 > 設(shè)計(jì)應(yīng)用 > 基于多特征融合的商品識(shí)圖匹配算法研究
基于多特征融合的商品識(shí)圖匹配算法研究
信息技術(shù)與網(wǎng)絡(luò)安全
王鑫城,范 紅,劉錫澤,胡晨熙,林 威,禹素萍
(東華大學(xué) 信息科學(xué)與技術(shù)學(xué)院,上海201620)
摘要: 隨著近幾年無(wú)人超市的不斷發(fā)展成熟,自助購(gòu)物越來(lái)越普及。如果商品售出后沒(méi)能及時(shí)補(bǔ)充,會(huì)影響消費(fèi)者的購(gòu)買(mǎi)意愿。為提升商品圖像識(shí)別的準(zhǔn)確率,采用多特征融合的方法,即將多種算法的特征聯(lián)合,形成優(yōu)勢(shì)互補(bǔ)。采用融合SIFT特征和灰度共生矩陣特征的方法完成貨架商品圖像的匹配。實(shí)驗(yàn)表明,該方法對(duì)比灰度共生矩陣方法準(zhǔn)確率提升20.6%,對(duì)比SIFT算法和PCA-SIFT算法準(zhǔn)確率分別提升8.9%和5.6%;處理時(shí)間對(duì)比以上三種算法略有增加。此方法還可用于分析貨物受歡迎程度以及確認(rèn)哪些柜臺(tái)需要加貨等,從而及時(shí)有效地對(duì)短缺的商品進(jìn)行補(bǔ)充。
中圖分類(lèi)號(hào): TP391
文獻(xiàn)標(biāo)識(shí)碼: A
DOI: 10.19358/j.issn.2096-5133.2021.04.011
引用格式: 王鑫城,范紅,劉錫澤,等. 基于多特征融合的商品識(shí)圖匹配算法研究[J].信息技術(shù)與網(wǎng)絡(luò)安全,2021,
40(4):70-74.
Matching algorithm of product image based on multi-feature fusion
Wang Xincheng,F(xiàn)an Hong,Liu Xize,Hu Chenxi,Lin Wei,Yu Suping
(College of Infrmation Science and Technology,Donghua University,Shanghai 201620,China)
Abstract: With the continuous development of unmanned supermarkets in recent years, self-service shopping has become more and more popular. If the goods are not replenished in time after they are sold, it will affect consumers′ willingness to buy. In order to improve the accuracy of product image recognition, a multi-feature fusion method is adopted, that is, the features of multiple algorithms are combined to form complementary advantages. The method of fusing SIFT features and gray level co-occurrence matrix features is used in this paper to complete the matching of shelf product images. Experiments show that the accuracy of this method is increased by 20.6% compared with the gray-level co-occurrence matrix method, and the accuracy of the SIFT algorithm and the PCA-SIFT algorithm are increased by 8.9% and 5.6% respectively; the processing time is slightly increased compared with the above three algorithms. This method can also be used to analyze the popularity of goods and confirm which counters need to be restocked, so as to supplement the shortage of goods in a timely and effective manner.
Key words : image matching;SIFT features;gray level co-occurrence matrix;feature fusion

0 引言

隨著人工智能的不斷發(fā)展,對(duì)圖像進(jìn)行特征提取并通過(guò)特征匹配完成對(duì)目標(biāo)的對(duì)比及識(shí)別,成為了計(jì)算機(jī)視覺(jué)領(lǐng)域中最主要的手段之一,并在大量領(lǐng)域有著廣泛的應(yīng)用,如人臉識(shí)別、車(chē)牌檢測(cè)、無(wú)人駕駛和醫(yī)學(xué)診斷等。對(duì)于商品貨架圖片而言,特征點(diǎn)的提取數(shù)量與匹配的精度會(huì)對(duì)商家后續(xù)的運(yùn)作有較大的影響。在圖像匹配算法研究中,最為常見(jiàn)的莫過(guò)于尺度不變特征變換算法(Scale-Invariant Feature Transform,SIFT),經(jīng)典SIFT算法由LOWE D G在1999年提出[1],并于2004年完善[2]。該算法穩(wěn)定性高,對(duì)旋轉(zhuǎn)、尺度縮放和亮度變換保持不變性。傳統(tǒng)SIFT算法的描述子是具有128維的特征向量,在特征點(diǎn)的特征向量生成以及最后進(jìn)行匹配的過(guò)程中需要大量的運(yùn)行時(shí)間。因此,降低特征向量的維數(shù)是學(xué)者們的一個(gè)重要研究方向,如SPCA、PCA[3]算法。另一個(gè)研究方向是采用特殊結(jié)構(gòu)[4],基于二分查找的思想,對(duì)得到的特征描述子進(jìn)行劃分,同時(shí)結(jié)合KNN算法,加快特征向量的匹配搜索速度,常用的結(jié)構(gòu)有KD-Tree[5]等。2006年BAY H提出SUFR(Speeded Up Robust Features)[6]算法,提升了特征提取速度,但在對(duì)尺度和旋轉(zhuǎn)的適應(yīng)性方面不及SIFT算法。秦緒佳[7]提出在特征匹配中引入灰度域和空間域的自相關(guān)性,提高匹配準(zhǔn)確性,但增加了算法復(fù)雜度,耗時(shí)大。RUBLEE E等[8]結(jié)合了BRIEF算法和FAST算法,提出了ORB算法來(lái)解決旋轉(zhuǎn)不變性問(wèn)題,但其特征符的區(qū)分性弱,匹配效果一般。



本文詳細(xì)內(nèi)容請(qǐng)下載:http://ihrv.cn/resource/share/2000003480




作者信息:

王鑫城,范  紅,劉錫澤,胡晨熙,林  威,禹素萍

(東華大學(xué) 信息科學(xué)與技術(shù)學(xué)院,上海201620)


此內(nèi)容為AET網(wǎng)站原創(chuàng),未經(jīng)授權(quán)禁止轉(zhuǎn)載。