In his report, Dean Qian Xiaojun introduced in detail the development status, existing problems, demand analysis and development trends of computer vision algorithms in the field of intelligent recognition.
Research Lecture on the Application of Computer Vision Algorithms in the Field of Intelligent Recognition
April 7, innovation expert of JMRH of Jiangsu Province, director of JMRH Key Laboratory of Three-dimensional Space Modeling and Simulation of Jiangsu Province, member of the Database Committee of the China Computer Society, deputy director of the Engineering Graphics Society of Jiangsu Province, Artificial Intelligence Research Institute of Nanjing Normal University Dean Qian Xiaojun was a guest at the "IIMT Intelligent Manufacturing Lecture Hall" and gave a special report titled"Research on the Application of Computer Vision Algorithms in the Field of Intelligent Identification". Nearly a hundred people from Jiangsu Jicui Intelligent Manufacturing Technology Research Institute's Innovation R&D Center, Smart Factory Center, Smart Software Center, and Humanoid Robot Company attended the meeting.
In his report, Dean Qian Xiaojun introduced in detail the development status, existing problems, demand analysis and development trends of computer vision algorithms in the field of intelligent recognition.
Dean Qian explained the five major tasks of computer vision to all R&D personnel of Jicui Intelligent Manufacturing and introduced in detail the target detection methods and applications based on computer vision, demonstrating the real-time, high-precision, simplicity and scalability of visual detection.
Through four cases of research on the detection method of dam disaster points based on dual-light images, research on the intelligent detection method of cofferdam collapse, research on the intelligent detection method of automobile wiring harness integrated final assembly, and the VR-based Gucheng Lake digital twin system, Dean Qian fully demonstrated the computer The application value of vision algorithms in different fields provides innovative solutions to real-life problems through a combination of theoretical research and practical operations.
Dean Qian said that the next step will be to optimize the efficiency of 3D space reconstruction and the technical points of video classification and recognition as the focus of work. Deep learning technology will be used to improve the performance of the 3D reconstruction algorithm; multiple subtasks will be divided and parallel computing methods will be used to accelerate processing. process; use efficient data compression algorithms to compress data, use filtering technology to remove noise and redundant information, and improve reconstruction efficiency; build high-quality and diverse data sets; optimize existing video classification and recognition algorithms and models; conduct in-depth research on how to Improve the accuracy of video classification and recognition in complex scenarios; explore the application of unsupervised or semi-supervised learning methods in video classification and recognition to improve the generalization ability of the model.
At the end of the report, some R&D personnel actively participated in the interaction.
In order to spread the culture of innovation, Jiangsu Jicui Intelligent Manufacturing Technology Research Institute has carefully created the "Jicui Intelligent Manufacturing Lecture Hall", inviting industry celebrities, experts and scholars to hold lectures to teach professional knowledge, industry cutting-edge, application prospects and other aspects to help industry, academia and research achieve high quality. develop. Welcome all walks of life to continue to pay attention to the "Collection of Intelligent Manufacturing Lecture Hall".
