High Resolution Visible Light Capable ANPR Internationalisation (HR-VL)

高分辨率可见光（HR-VL）自动车牌识别（ANPR）国际化项目将研究新的图像处理和高清数字cctv摄像机控制算法，以采用基于英国的新方法，使ANPR进入新的国际市场。号牌形式，格式和位置上的车辆在英国和欧洲是很受限制的。世界上许多其他地方的情况并非如此。例如，在美国，车牌的背景格式可以由图像、可变颜色等组成。这使得ANPR相当困难。研究/创新的关键领域将是：a)国际化——使图像处理算法能够处理非英国车牌。这需要处理广泛的字体/字符集（对阿拉伯语、西里尔文和中文更感兴趣），处理车牌背景，同时补偿车牌表面的典型缺陷；b)即插即用-使安装尽可能简单，除实际安装访问外，无需用户出席。这就需要新的闭路电视摄像机控制软件作为国际化的必要先导。对新算法的成功研究和部署将：a)使北美、亚洲和中东的新国际市场成为可能；b)降低部署和维护成本。自配置高清数字闭路电视摄像机可以使用更少的摄像机来覆盖每个道路配置；c)使第三方供应商能够创建和部署一系列新的监控产品和服务。这包括自动停车场付款、危险材料追踪和移动交通违规行为。这个为期9个月的概念验证项目的关键成果是通过在演示器中的部署和贬值来评估图像科学和高清CCTV摄像机控制算法的新集合。

The High Resolution Visible Light (HR-VL) Capable Automatic Number Plate Recognition(ANPR) Internationalisation Project will research new image processing and HD DigitalCCTV camera control algorithms required to take a novel UK-based approach for ANPR intonew international markets. Number plate forms, formats and location on vehicles in the UKand Europe are well constrained. This is not the case in many other parts of the World. Forexample, in the USA, the background format for a number plate can consist of images,variable colouring, etc. This makes ANPR considerably more difficult. The key areas ofresearch/innovation will be:a) Internationalisation – to enable the image processing algorithms to handle non-UK numberplates. This requires processing a wide range of fonts/character sets (with additional interest inArabic, Cyrillic & Chinese), processing of plate backgrounds while compensating for typicaldefects on the number plate surface;b) Plug-and-play – making installation as simple as possible and removing the need for otherthan the actual installation visit to require user attendance. This requires new control softwarefor the CCTV camera as a necessary precursor to internationalisation.Successful research into, and deployment of, new algorithms will:a) Enable new international markets in North America, Asia and the Middle East to becomeavailable;b) Reduce the cost of deployment and maintenance. Self-configuring HD Digital CCTVcameras enable a deployment to use fewer cameras to cover each road configuration;c) Enable a new range of surveillance products and services to be created and deployed bythird party vendors. This includes automated car park payment, hazardous materials trackingand moving traffic violations.The key deliverable of this nine-month Proof-of-Concept Project is the evaluation of the newset of image science and HD CCTV camera control algorithms through their deployment andevaluation in a demonstrator.