Design of a user-support system for vision information using a smart phone

摘要

In this paper, we design a user-support system for vision information using smart phones. When the user takes a picture of a target using the smart phone camera, our system can recognize the image of the target and recommend a suitable service to the user. The system executes simple image analysis and determines a suitable service for the target image. The simple image analysis can extract 13 parameters (e.g., color information (RGB), number of intersections, depth of intersection, line width, and line depth). We use machine learning to classify the parameters into each service. In our research, we use a Support Vector Machine as a learning machine tool. Our system design provides user support for several services such as translation, barcode readers, and diagnosis of skin images and demonstrates the effectiveness of our research.