Within the area of environmental perception, automatic navigation, object detection, and computer vision are crucial and demanding fields with many applications in modern industries, such as multi-target long-term visual tracking in automated production, defect detection, and driverless robotic vehicles. The performance of computer vision has greatly improved recently thanks to developments in deep learning algorithms and hardware computing capabilities, which have spawned the creation of a large number of related applications. At the same time, with the rapid increase in autonomous systems in the market, energy consumption has become an increasingly critical issue in computer vision and SLAM (Simultaneous Localization and Mapping) algorithms. This paper presents the results of a detailed review of over 100 papers published over the course of two decades (1999–2024), with a primary focus on the technical advancement in computer vision. To elucidate the foundational principles, an examination of typical visual algorithms based on traditional correlation filtering was initially conducted. Subsequently, a comprehensive overview of the state-of-the-art advancements in deep learning-based computer vision techniques was compiled. Furthermore, a comparative analysis of conventional and novel algorithms was undertaken to discuss the future trends and directions of computer vision. Lastly, the feasibility of employing visual SLAM algorithms in the context of autonomous vehicles was explored. Additionally, in the context of intelligent robots for low-carbon, unmanned factories, we discussed model optimization techniques such as pruning and quantization, highlighting their importance in enhancing energy efficiency. We conducted a comprehensive comparison of the performance and energy consumption of various computer vision algorithms, with a detailed exploration of how to balance these factors and a discussion of potential future development trends.

This paper introduces an optimization method for multi-robot automated control welding based on a Particle Swarm Genetic Algorithm (PSGA), aiming to address issues such as high costs, large footprint, and excessive production cycles in multi-robot welding production lines. The method first constructs a multi-axis robotic kinematic model to provide constraint conditions. Then, the PSO (particle swarm optimization) algorithm, which integrates penalty functions into the fitness evaluation, is used to determine the optimal welding path by simulating collective behavior within a group. The GA (genetic algorithm) encodes the position of the welding robot bases into chromosomes to find the optimal layout for coordinated control of multiple robots. The entire process is optimized according to welding standards and requirements. Additionally, a comprehensive production line performance estimation model was used to quantitatively analyze the new scheme. The results show that the optimized production line’s balance rate increased by 10%, the balance loss rate decreased by 10%, the smoothness index increased by 37.8%, the space costs reduced by 44.4%, the equipment demand reduced by 41.1%, the labor demand reduced by 50%, the total costs reduced by 10%, and the average product cycle time was reduced by 5.07 s. Finally, we tested the algorithm in various complex scenarios and compared its performance against mainstream algorithms within the context of this study. The results demonstrated that the optimized production line significantly improved efficiency while maintaining safety standards.