关键词: 图像识别；EasyDL；同步定位与建图；增强现实车书

Abstract: At present, there are some problems with complicated text content, difficult retrieval, strong professionalism, difficult understanding and low usage for vehicle user manual. To solve the above problems, this paper proposes an image recognition method based on the EasyDL deep learning, and then the recognized icon information is located in real-time on the icon side through visual simultaneous localization and mapping (SLAM). This method can be used to recognize the characters, geometry, and parts of the vehicle accurately, and map the recognized content with a three-dimensional way on interactive devices. The theoretical analysis and experimental results show that the method described in this paper has the characteristics of fast calculation speed, high recognition efficiency, intuitive display, easy to understand, etc., and has important practical value.

Key words: Image recognition; EasyDL; Simultaneous localization and mapping; Augmented reality vehicle manual