Author: Zen and the Art of Computer Programming
1 Introduction
With the development of emerging technologies such as Industrial IoT (Industrial IoT), artificial intelligence (AI), robotics, and biometric technology, in-depth changes have also occurred in the field of industrial automation. The development of industrial automation will drive the development of computer graphics and deep learning. This article starts with visual tasks such as image classification, target detection, image segmentation and depth estimation, and explains the related applications of computer graphics and deep learning in industrial automation.
1. Background introduction
Industrial automation mainly consists of mechanical, electrical, control, process control, electronics and management fields. Among them, computer vision and deep learning are a major focus area for the development of industrial automation. Under the influence of the industrial Internet of Things, artificial intelligence, robotics, and biometric technology, computer vision and deep learning in the field of industrial automation are booming. The "Action Plan for the Development of Industrial Automation during the 13th Five-Year Plan" promulgated by the China Enterprise Manufacturing Association in the first quarter of 2019 clearly proposes to promote the application of computer vision and deep learning in the field of industrial automation, and to promote computer vision and deep learning in the field of domestic industrial automation. Learn research to conduct investigation, evaluation and planning, including industrial control, mechanical engineering, ecological monitoring, smart agriculture and other fields. The "Action Plan for the Development of Industrial Automation during the "Thirteenth Five-Year Plan"" also predicts that by 2020, the overall application of computer vision and deep learning in the field of industrial automation will rise to more than 80%. At the same time, big data, blockchain and other emerging technologies will become important technological breakthroughs in the field of industrial automation. In this article, I will start from several typical cases of visual tasks and introduce how to use computer graphics and deep learning to solve these visual tasks. First, I will introduce the basic concepts and terminology of computer graphics, and then use simple cases to introduce some applications of computer graphics, such as image classification and target detection; then, I will introduce the basic concepts and terminology of deep learning, and discuss how to Train a deep neural network to achieve image segmentation; finally, I will show readers how to combine these two techniques to achieve a more complex target detection task-target tracking.