Express Package Sorting System Based On Printed Chinese Character Recognition

In recent years, a large number of automatic sorting system based on bar code and two-dimensional code has been applied to the sorting process with the rapid development of the express service industry and the continuous improvement of information technology. However, the situation of barcode defaced happens from time to time for the complexity of the express sorting environment. For the package with unreadable barcode,courier companies usually use artificial identification of send and receive address characters to achieve sorting, which has high cost and low efficiency. Chinese character recognition technology based on image processing has the advantages of high speed, low cost and high degree of automation, which has a broad application prospect in express parcel sorting system.Firstly, the overall structure and working principle of the sorting system is introduced, and the design and selection of hardware platform and the design of software platform. Then, express bill location is obtained by using multi thresh Otsu method,morphological operation and connected domain selection according to the characteristics of high brightness, and a supplementary positioning scheme is designed by using the high contrast feature of the barcode area. Then, the angle of the image is corrected by Hough transform, and the position of the character is found after multiple analyzing the projection waveform. And the character image is split and normalized. In character recognition phase, the production process of the sample character database is described firstly, the feature extraction process of the character image by using the feature of pixel grid and gradient direction grid is described secondly. Finally, the character recognition is achieved by to the standard Euclidean distance classifier.The development outcomes of hardware environment and software platform of the parcel sorting system is presented, and it is verified by a large number of packages. The results show that the character recognition rate reached 99.76%, parcel sorting accuracy rate reached 98.35% and the average processing time is 0.5 seconds, which meets the requirements of express sorting system. this system can be used to reduce labor costs,improve the efficiency of sorting, or combined with the barcode recognition technology to improve the redundancy of the sorting system.