Welding spot defect detection system based on computer vision
A technology of computer vision and defect detection, applied in computing, image data processing, instruments, etc., can solve the problems of extracting too many features and increasing computational complexity, and achieve the effect of improving detection speed
Patent Information
- Authority / Receiving Office
- CN · China
- Current Assignee / Owner
- Publication Date
- 2018-01-19
- Estimated Expiration
- Not applicable · inactive patent
Smart Images

Figure 1
Abstract
Description
Technical field
[0001] The invention relates to a detection system, in particular to a computer vision-based solder joint defect detection system. Background technique
[0002] In recent years, the detection technology based on computer vision has developed rapidly in the field of industrial detection. It combines modern science such as modern optics, computer technology, image processing and analysis technology into a comprehensive detection system of optomechanical integration, which has online detection and real-time analysis. And control, high precision, continuous work and other characteristics, can be widely used in various special occasions. The spot welding machine is mainly used in the equipment of the electronics industry, which has high quality requirements. Therefore, the quality of the solder joints needs to be tested. Due to the immaturity of the actual process and errors in manual operation, some defects often occur, such as too much solder, too little solder...
Examples
Embodiment Construction
[0018] The solder joint defect detection system based on computer vision is mainly composed of video acquisition equipment, control unit, PC and PC detection software. The video acquisition equipment is composed of industrial digital cameras, optical lenses and light sources. The industrial digital camera and lens are used for image acquisition of the measured object, and are connected to the host computer through the 1394 interface, and the collected image data is transmitted to the host computer for processing. In order to meet the detection accuracy of 0.1 mm, this system uses a 5 million-pixel industrial surface-scanning CCD, so that the size of a single pixel is 0.04 mm. The light source is used to illuminate and highlight the characteristics of the object. The system uses a coaxial white light source of LED particles, and adopts a front lighting method to make the background brightness uniform, that is, the pixel gray value remains consistent. The upper computer is the c...