Briefly describe the limitations of machine vision technology in industrial manufacturing

  A typical machine vision application system includes image capture module, image digitization module, image processing module, decision-making module, mechanical control execution module and light source system. So what are the limitations of the application of machine vision technology in industrial manufacturing? Let’s learn about it together:

1. Constrained by environmental light sources: Different light sources will cause different imaging quality and effects, directly interfere with the detection of the detection algorithm, and may cause misjudgment of the product; a single visual guidance technology cannot guarantee the accuracy of obstacle detection in the path, decision-making The control layer often needs to fuse information collected by multiple sensors.

2. Restricted by the performance of hardware equipment: lens distortion correction, calibration differences, and limited viewing angle range of the camera; installation conditions and site restrictions, and requirements for sensor fusion solutions; each pixel has different dark current and inconsistent response to photons, which will cause Causes spatial and pattern noise in the camera; limitations of parameter settings of CCD line array cameras.

3. Limited by on-end computing resources: The large-scale and complex model architecture of industrial products requires strong computing power. When the memory on the device terminal is insufficient, the model needs to be trained offline in the cloud and then deployed to the device terminal; image data transmission still requires Adjusting and optimizing model parameters for specific mission objectives will generate additional engineering overhead and have poor real-time performance.

4. Limited by the diversity of detection objects: there are many types of defects on the surface of objects, the defect generation mechanism is unknown, and the description of defects is insufficient; it is difficult for the machine vision system to extract features from the data.

5. Restricted by cost and revenue economics: The development of core components of industrial cameras such as vision sensors and underlying vision software requires large investment costs.