| dc.description.abstract | In modern logistic company, Automated Sortation System as ASS (figure 0) offer a highly accurate and efficient means of automatically sorting, routing, a wide range of parcel type to unique destinations for the purpose of selection, processing, storing and shipping. However in ASS, to make sure the successfully of sortation, the system must recognize the parcel on the transfer line, secondly recognized the parcel’s document in order to get the document information which relate to optical character detection and regconiton in digital image processing. So far, there has been amount of research of optical character recognition. There are amount of research about this usually about car plate regconition, motobike plate regconition, handwriting regconition (figure 1). But apply to ASS, to get the document information these studies mostly come to problems that the parcel detection and parcel document recognition are hard due to parcel condition, document condition, light, human interaction. In this thesis will present a method which mainly focus on detecting Vietnam Shopee parcel’s document and extracting its information. The proposed method conntent two parts are detecting the ShopEE document on parcel using Opencv Object Detection (OOD)_ Canny and Contour method and extracting the information as much as possible of the parcel’s document using Optical Character Recognition (OCR) method. The inputs are the image file that have ShopEE parcel’s document. Some experiments will be conducted to give the best result method for parcel’s document detection and extracting its information to txt file. | en_US |
