自动识别与定位服务在书库管理中的实现与研究

发布时间：2018-05-03 06:29

本文选题：条码识别 + 室内定位　；参考：《南京邮电大学》2017年硕士论文

【摘要】：物流仓储作为物流行业最关键的环节之一,其自动化和智能水平直接影响整个物流服务的质量和效率。目前,物流仓储处于自动化水平偏低的发展阶段,同时还存在一些问题,例如书库管理中的图书借阅和上架流程,包括查询图书标签,确定书库位置,查询货架编号,寻找图书货架等,其流程繁琐,效率低,自动化水平不高,尤其对于不熟悉书库环境的来访者来说,从查询所需书信息到获取的过程相当复杂,不能快速获取图书位置及寻书路线。为了满足书库管理中的这一需求,本文设计并实现了自动识别与定位服务系统,本文做了如下研究工作:(1)提出了动态梯度匹配K-近邻算法,对于动态Wi-Fi环境中的室内定位精度大大下降且性能不稳定,数据库中数据冗余导致的匹配效率低等问题,该算法展示了在动态Wi-Fi环境中Wi-Fi定位的抗干扰性,同时提出的特征指纹的梯度化和二级分区分别减少了数据库冗余和提高了分区的命中率。实验结果证明,该算法在整体定位精度、定位效率和分区命中率有明显提高以及设备的异质性有明显的优势。(2)提出了动态联合跟踪与导航机制,对于Wi-Fi盲区环境中Wi-Fi定位的失效,对当前位置和目的地之间的路线规划问题,该机制实现了Wi-Fi盲区环境中实现持续室内定位。同时,提出了基于参考指纹节点的实时动态Dijkstra导航算法,实现了两个位置之间的最优路径规划。实验结果证明,该算法在路径规划长度和时间使用上有明显提高。(3)本文设计的自动识别与定位服务系统并在书库管理中实现,该系统能很好的解决了书库管理中寻书和图书上架效率低、查询效率慢、流程繁琐的特点和问题,成功地为图书仓库中管理员和来访者提供可靠的图书信息查询、室内定位和导航的实用性工具。
[Abstract]:Logistics warehousing as one of the most critical links in logistics industry, its automation and intelligence level directly affect the quality and efficiency of the entire logistics services. At present, logistics warehousing is on the low level of automation, and there are still some problems, such as the process of book borrowing and loading in library management, including querying the label of books, determining the location of the library, and querying the shelf number. The process of searching for book shelves is cumbersome, inefficient, and low level of automation. Especially for visitors who are not familiar with the library environment, the process from querying the required book information to obtaining it is quite complex. Unable to quickly obtain book location and book seeking route. In order to meet the requirement of library management, this paper designs and implements an automatic recognition and location service system. In this paper, the following research work is done: 1) A dynamic gradient matching K-nearest neighbor algorithm is proposed. For the problems of low precision and unstable performance of indoor positioning in dynamic Wi-Fi environment and low matching efficiency caused by data redundancy in database, the algorithm shows the anti-interference property of Wi-Fi location in dynamic Wi-Fi environment. At the same time, the gradient and two-level partitioning of the feature fingerprint can reduce the redundancy of database and improve the hit ratio of the partition. The experimental results show that the algorithm has obvious advantages in overall positioning accuracy, location efficiency and area hit ratio, as well as heterogeneity of equipment. (2) A dynamic joint tracking and navigation mechanism is proposed. For the failure of Wi-Fi location in Wi-Fi blind area environment and the route planning problem between current location and destination, the mechanism realizes continuous indoor positioning in Wi-Fi blind area environment. At the same time, a real-time dynamic Dijkstra navigation algorithm based on reference fingerprint nodes is proposed to realize the optimal path planning between two locations. Experimental results show that the algorithm has a significant increase in the length and time of path planning.) the automatic identification and location service system designed in this paper is implemented in the library management. The system can solve the problems of low efficiency of searching books and bookshelves, slow query efficiency, cumbersome process, and successfully provide reliable book information query for librarians and visitors in the library management. Practical tools for indoor positioning and navigation.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：G250.7;TP391.4;TN92

【参考文献】