Industrial machine vision system architecture is mainly divided into two parts: hardware equipment and software algorithm. The hardware equipment mainly includes light source system, lens, camera, image acquisition card and vision processor; the core algorithm in the software package mainly includes traditional digital image processing algorithm and image processing algorithm based on deep learning.
The visual sensor in the industrial camera lens is continuously optimized in structural design;
The application of machine vision system increases the efficiency of industrial field programming;
The device-side deep learning model is continuously compressed and accelerated;
The improvement of computing power on the device side;
Combining computer vision and robotics technology to increase robot vision adaptive capabilities.
Micron-level non-destructive testing can be performed on defects of 3D printed products;
Visual information enhances the ability of autonomous perception in the processing of intelligent machine tools;
The application of intelligent vision equipment improves the operating efficiency and safety of factory employees;
Let industrial robots learn vision-based motor skills and operation strategies from actual work;
Automatic evaluation of the image quality of the cell microscope in the cytology research work.