Research and Develop Smartwatch Display Recognition System

Abstract

Image recognition (or image classification) is the task of identifying images and categorizing them in one of several predefined distinct classes. Therefore, image recognition software and apps can define what is depicted in a picture and distinguish one object from another. The field of study aimed at enabling machines with this ability is called computer vision. Being one of the computer vision (CV) tasks, image classification serves as the foundation for solving different CV problems [1]. To realize the recognition of images, the first to get corresponding image by image acquisition device, so that the digital image; Then the image recognition, and its various information. In this research, a neural network is used to analyze the acquired digital image recognition, the neural network is introduced into the image recognition field, and combined with conventional digital image processing technology, find out a kind of strong accuracy plane image recognition method. The deep neural network has been trained and tested with a self-built database consisting of icon images in different conditions. The system with a new developed vision module requires an appropriate development kit to execute the vision tasks and a suitable protocol scheme. To build a recognition system including the image recognition task, we use the Nvidia Jetson TX2 developer kit. Apparently, the obtained results show that the proposed system is intensely effective with the accuracy reaching to 92%. In addition, the protocols in the hardware part have been selected appropriately to optimize efficiency of several important sub-tasks. Further analysis among examined algorithms have been presented in this study.