车牌号识别技术是计算机视频图像识别技术在车辆牌照识别中的一种应用。通过手机摄像头“扫一扫”的方式，可以在1秒内识别出车牌号码，精确度达99%，支持包括蓝牌、黄牌（双）、军牌（双）、武警牌（双）、警牌、农用车牌、教练车牌、港口车牌、大使馆等各种规格牌照的识别；车牌宽度要求低，车牌单字符高度18个像素，都不影响识别；车牌号识别核心纯C代码编写，可以跨平台应用，适用于汽修、车辆投保理赔、移动警务、智能安防、汽车金融、汽车后市场、移动接车等服务。

License plate recognition technology is an application of computer video image recognition technology in vehicle license plate recognition. By "scanning" the mobile phone camera, the license plate number can be identified within 1 second, with an accuracy of 99%. It supports the identification of various specifications of license plates, including blue plate, yellow plate (double), military plate (double), armed police plate (double), police plate, agricultural plate, coach plate, port plate, embassy plate, etc. The width requirement of license plate is low, and the height of single character of license plate is 18 pixels, which does not affect the recognition; License plate identification core pure C code writing, can be cross-platform application, suitable for auto repair, vehicle insurance claims, mobile police, intelligent security, auto finance, auto aftermarket, mobile car pickup and other services.

车牌号识别技术主要分为：车牌定位、车牌提取、字符分割、字符识别。车牌号识别实现的方式主要分为两种：一种是静态图像图片的识别，另一种是动态视频流的实时识别。静态图像识别技术的识别有效率较大程度上受限于图像的抓拍质量，为单帧图像识别；而动态视频流的识别技术适应性较强，识别速度快，它实现了对视频每一帧图像进行识别，增加识别比对次数，择优选取车牌号，关键在于较少的受到单帧图像质量的影响。

License plate recognition technology is mainly divided into: license plate location, license plate extraction, character segmentation, character recognition. There are two ways to realize license plate recognition: one is the recognition of static image, the other is the real-time recognition of dynamic video stream. The recognition efficiency of static image recognition technology is limited by the quality of image capture to a large extent, which is single frame image recognition. The recognition technology of dynamic video stream has strong adaptability and fast recognition speed. It realizes the recognition of every frame of video image, increases the number of identification comparison, and selects the license plate number optimally. The key is that it is less affected by the quality of single frame image.