本课题以西安科技大学自动物流系统为平台,设计并制作了基于数字图像条码识别技术的货物分拣系统。针对传统条码识别器受读取距离和条码形变等因素影响较大的缺点，提出一种基于决策树和AdaBoost算法融合的图像条码识别方法。通过Matlab对算法进行仿真验证，进而在ARM Cortex M3上实现了对条码图像的采集、处理和识别。实验证明，该算法在容错性、响应速度和对条码识别的通用性都有很大提升。针对现有的物流系统在运行过程中不能自动分拣货物，设计了一种基于数字图像条码识别技术的货物分拣系统，该系统能根据来货的条码为货物分配库位，并记录入库货物的数量，借助短信可以全天候向仓库管理人员反馈仓库货物存储信息。分拣系统的加入，极大方便了仓库管理和运营。 本论文介绍和分析了本校的自动物流系统的各个结构模块，介绍了物流系统的运行原理，进而阐述了加入自动分拣系统的必要性。接着分析了自动分检系统的可行性，提出了实现该分检系统的软件和硬件可行方案。 本课题以Code39码为例，分析了该条码的编码结构，根据其特点建立了特征向量空间。提出了基于改进AdaBoost算法的条码识别方法，并将其应用于货物智能分拣系统中。通过Matlab对改进后的AdaBoost算法对条码的识别效果进行了验证。 在验证基于改进AdaBoost算法的货物分拣方法的可行性后，将其移植入以ARM Cortex M3为核心的硬件平台中，并运用uC/OSII实时操作系统调度各个功能模块的动作，通过调试和实验检验了算法的效果。为了能够让自动分拣系统全天候向用户反馈仓库信息，分检系统加入了短信收发模块，并定义了相关的通信协议帧，使得用户能在任何有手机信号覆盖的地方查询库存。 经过大量的实验和测试，该系统能够准确识别和分拣货物并能记录每种货物的库存。同时可以响应任意一种手机的短信查询命令，并向用户返回当前库存的准确信息，为仓库的管理和决策提供保障。

A warehouse intelligent sorting system is been designed using digital image recongnition technology based on the platform of automated logistics system of XUST (Xi'an University of Science and Technology) in this thesis. Aming at the drawback of traditional barcode reader vulnerable to the effects on read distance and deformation, a barcode image recognition method based on the integration of decision trees and improved AdaBoost algorithm is proposed. Through matlab simulation and transplant the method into ARM Cortex-M3 platform, the function of barcode image acquisition, processing and recognition are implemented. Experimental results show that the algorithm has a signigicant enhancement in fault tolerance, response speed and versatility barcode recongnition. For the existing logistics system can not sorting goods automaticly, an auto sorting system based on digital barcode image recongnition technology is designed. The system can distribute cargo location of warehose according to the goods barcode and can record the number of inbound shipment. With SMS, the system can feedback goods store information to warehouse managers around the clock. This auto sorting system brings great convenience for warehouse management and operations. The hardware and the operating principle of the automatic logistics system in XUST is introduced and analysed, which explains the need for automatic sorting system. Then the feasibility of automatic sorting system is analyzed, and the feasible options of both software and hardware for the sorting system are given. Take the example of Code39, the structure of the barcode is analyzed and feature vector space is constructed according the features of the barcode. A barcode recognition method based on improved AdaBoost algorithm is proposed, and applied it to the cargo intelligent sorting system. Through matlab simulation, the barcode recognition result of the improved AdaBoost algorithm has been verified. After the verification of the cargo sorting method based on improved AdaBoost algorithm, the method is transplanted into the ARM Cortex-M3 hardwre platform, and uC/OSII RTOS is used to schedule the various function modules. Through debugging and experiments, the effect of the algorithm has been tested. In order to feedback store information of warehouse around the clock, SMS transceiver is added in the smart sorting system, and the communication protocol is defined so that the managers can check inventory everywhere. After a lot of experiments and tests, this system is able to accurately indentify and sorting goods and can recod each goods inventory. Also the system can respond to SMS query command of any mobile phone, and feedback accurate information of current inventory of the warehorse. This system provides guarantee for warehouse management and decision-making.