Research Progress on the Integration of Robot Vision, Computer Vision and Machine Learning: Technological Evolution, Challenges and Industrial Applications

Abstract

This paper systematically reviews the technological progress of the integration of robot vision, computer vision (CV) and machine learning (ML), focusing on the design paradigms of global vision system, local embedded vision, hybrid cloud edge architecture, as well as target detection, 3D reconstruction, Algorithm innovation of CV technology such as dynamic scene understanding.Combining the seven major trends of IDC 2025 embodied intelligent robots with the China machine vision market research report, this paper analyzes solutions to challenges such as real-time, data scarcity and multi-modal fusion and proposes solutions based on lightweight models, federated learning and neural symbolic systems.Future direction.By citing top conference papers and industry white papers such as CVPR, ICRA and NeurIPS, this paper builds a technology-scenarioindustry closed-loop academic framework to provide theoretical support for the intelligent upgrade of robot vision.In recent years, with the rapid development of artificial intelligence and computer vision technology, robot vision system has made remarkable progress.These systems improve the perception and decision-making capabilities of robots in complex environments by combining deep learning and machine vision technologies.Deep learning algorithms excel in image processing and feature extraction, enabling robots to more accurately identify and track target objects1-5.In addition, the progress of machine vision technology also provides strong support for the application of robots in agriculture, industry, medical and other fields6-10.In the agricultural field, machine vision technology is widely used in tasks such as robot navigation and fruit detection and efficient automated operation is achieved by optimizing the robot control system through deep learning algorithms6,7,11.In the industrial field, robot vision system improves the accuracy and efficiency of welding, sorting and other tasks through the combination of deep learning and machine vision technology10,12,13.In addition, in special scenarios such as medical treatment and fire protection, the robot vision system also shows its unique advantages.Through deep learning and visual servo technology, the robot can learn and adapt to complex environments autonomously14,15.