【摘要】：在人們的日常生活中經(jīng)常會(huì)使用到各種表單或者票據(jù),他們是很重要的法律憑證。為了自動(dòng)對(duì)表單或者票據(jù)中的簽名驗(yàn)明真?zhèn)?比如銀行存取款憑證簽名真?zhèn)巫R(shí)別、保險(xiǎn)公司保單簽名真?zhèn)巫R(shí)別、快遞單據(jù)識(shí)別等等,需要首先確定簽名在表單中的位置。現(xiàn)如今大部分表單的簽名真?zhèn)巫R(shí)別工作都是由人工來(lái)完成的,費(fèi)時(shí)費(fèi)力且受人為主觀因素影響較大。因此,開(kāi)發(fā)一套表單手寫(xiě)簽名自動(dòng)檢測(cè)系統(tǒng)具有十分重要的研究意義。本文主要研究了手寫(xiě)簽名自動(dòng)檢測(cè)系統(tǒng)中的目標(biāo)定位功能。目標(biāo)定位一直是一個(gè)熱點(diǎn)的問(wèn)題,目前常用的方法有基于顏色、紋理、形狀、空間、模板匹配等方法。本文采用卷積神經(jīng)網(wǎng)絡(luò)獲取候選區(qū)域的方法,實(shí)現(xiàn)了表單中手寫(xiě)簽名的定位功能。首先,使用卷積神經(jīng)網(wǎng)絡(luò)獲取候選區(qū)域,其實(shí)質(zhì)是使用滑動(dòng)窗口對(duì)圖像進(jìn)行窮舉搜索;其次,利用分類(lèi)層對(duì)每個(gè)特征圖上的候選區(qū)域做分類(lèi)任務(wù),判斷該候選框是前景還是背景;最后,利用回歸層對(duì)每一個(gè)候選框的位置做回歸。本文主要研究?jī)?nèi)容有:(1)收集了大量、多樣的表單圖像,自定義表單圖像數(shù)據(jù)庫(kù)。收集的圖像包含銀行表單、保險(xiǎn)公司表單、快遞公司表單等,對(duì)收集好的表單圖像進(jìn)行旋轉(zhuǎn)、平移、縮放、加噪等處理,這樣設(shè)計(jì)出的表單圖像數(shù)據(jù)庫(kù)會(huì)使得訓(xùn)練結(jié)果更好。(2)為了提高區(qū)域候選框質(zhì)量,根據(jù)設(shè)計(jì)的表單圖像數(shù)據(jù)集中目標(biāo)的長(zhǎng)寬比,優(yōu)化滑動(dòng)窗口中獲取候選框的窗口比例。(3)為了減少高層卷積細(xì)節(jié)特征信息損失,盡可能多的獲取到更多圖像特征,對(duì)RPN網(wǎng)絡(luò)模型連接結(jié)構(gòu)進(jìn)行改進(jìn)。文中提出了 RPN-X的網(wǎng)絡(luò)結(jié)構(gòu),并且對(duì)該網(wǎng)絡(luò)模型中的訓(xùn)練參數(shù)進(jìn)行了優(yōu)化。(4)基于caffe框架,搭建訓(xùn)練神經(jīng)網(wǎng)絡(luò)的平臺(tái),并訓(xùn)練出RPN-X神經(jīng)網(wǎng)絡(luò)模型。(5)在matlab編程環(huán)境下,開(kāi)發(fā)了一個(gè)手寫(xiě)簽名位置定位系統(tǒng),實(shí)現(xiàn)了實(shí)時(shí)拍照上傳并且進(jìn)行手寫(xiě)簽名位置的定位。
[Abstract]:Forms or notes are often used in people's daily lives, and they are important legal documents. In order to automatically verify the authenticity of the signature in the form or bill, such as bank deposit and withdrawal certificate signature authenticity identification, insurance company policy signature authenticity identification, express document identification and so on, it is necessary to first determine the position of the signature in the form. Nowadays, the identification of signature authenticity of most forms is done manually, which is time consuming and influenced by human subjective factors. Therefore, the development of a form handwritten signature automatic detection system has a very important significance. In this paper, the function of target location in automatic detection system of handwritten signature is studied. Target location has always been a hot issue. At present, the commonly used methods are based on color, texture, shape, space, template matching and so on. In this paper, we use convolutional neural network to obtain candidate regions and realize the localization function of handwritten signature in the form. Firstly, using convolutional neural network to obtain candidate regions, the essence of which is to search the images by sliding window. Secondly, the candidate regions on each feature map are classified by classification layer. Finally, the location of each candidate is regressed by regression layer. The main contents of this paper are as follows: (1) collect a large number of various form images and customize the form image database. The collected images include bank forms, insurance company forms, courier company forms, etc. The collected images are rotated, translated, scaled, noised, etc. The designed form image database will make the training result better. (2) in order to improve the quality of the region candidate, according to the aspect ratio of the target in the designed form image dataset, (3) in order to reduce the loss of high-level convolution detail feature information and obtain as many image features as possible, the RPN network model connection structure is improved. In this paper, the network structure of RPN-X is proposed, and the training parameters in the network model are optimized. (4) based on the caffe framework, the platform of training neural network is built, and the RPN-X neural network model is trained. (5) in the matlab programming environment, A position location system for handwritten signature is developed, which can locate the location of handwritten signature by taking pictures and uploading them in real time.
【學(xué)位授予單位】：西安理工大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類(lèi)號(hào)】：TP391.41;TP183

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 余殷博;;基于人工智能下的機(jī)器學(xué)習(xí)歷史及展望[J];電子技術(shù)與軟件工程;2017年04期

2 顏飛;周長(zhǎng)久;田彥濤;;用于目標(biāo)定位的圖像邊緣點(diǎn)檢測(cè)算法[J];吉林大學(xué)學(xué)報(bào)(工學(xué)版);2016年06期

3 肖雄;黃樟燦;;基于統(tǒng)計(jì)的車(chē)牌字符識(shí)別[J];數(shù)字技術(shù)與應(yīng)用;2016年04期

4 李岳云;許悅雷;馬時(shí)平;史鶴歡;;深度卷積神經(jīng)網(wǎng)絡(luò)的顯著性檢測(cè)[J];中國(guó)圖象圖形學(xué)報(bào);2016年01期

5 盧宏濤;張秦川;;深度卷積神經(jīng)網(wǎng)絡(luò)在計(jì)算機(jī)視覺(jué)中的應(yīng)用研究綜述[J];數(shù)據(jù)采集與處理;2016年01期

6 閻沖;;基于SIFT算法的目標(biāo)特征檢測(cè)與提取技術(shù)研究[J];傳感器世界;2012年09期

7 楊新鋒;;圖像文本識(shí)別中目標(biāo)定位方法研究[J];微型電腦應(yīng)用;2012年05期

8 甄巍松;李國(guó)強(qiáng);魯統(tǒng)偉;;基于特征點(diǎn)相似度的匹配定位算法[J];武漢工程大學(xué)學(xué)報(bào);2011年04期

9 李亞標(biāo);王寶光;李溫溫;;基于小波變換的圖像紋理特征提取方法及其應(yīng)用[J];傳感技術(shù)學(xué)報(bào);2009年09期

10 賈小軍;喻擎蒼;;基于開(kāi)源計(jì)算機(jī)視覺(jué)庫(kù)OpenCV的圖像處理[J];計(jì)算機(jī)應(yīng)用與軟件;2008年04期

相關(guān)博士學(xué)位論文 前1條

1 鄧集杰;支票印鑒快速檢測(cè)方法中的關(guān)鍵技術(shù)研究[D];天津大學(xué);2010年

相關(guān)碩士學(xué)位論文 前2條

1 何柳;表單識(shí)別中的關(guān)鍵問(wèn)題研究[D];沈陽(yáng)工業(yè)大學(xué);2016年

2 章?tīng)T明;基于形狀特征的目標(biāo)檢測(cè)算法研究[D];華南理工大學(xué);2013年

，

本文編號(hào)：2258789