| The horizontal arm, which connects the body and the drive system, is an important part of the auto and directly related to the vehicle's performance. The weld quality of the horizontal arm is the most important one. There are 16 CO2 gas shielded arc weld seams on the whole work piece. Most of the weld seams are fillet welds in the shape of space curve. They are obtained by the industrial robot welding and may have partial welding, weld missing, surface pores and other defects.Currently, the weld quality of the horizontal arm is inspected by the worker's eyes. The manual weld quality inspection of auto parts is inefficient, low accuracy, and not easy to deal with data. An automatic system based on machine vision is developed to overcome the drawbacks. The system uses a robot arm to hold the CCD and the light sources in order to acquire the structured light images and the LED images of the weld seams. Adaptive image processing algorithms are designed according to the different image features. The missing weld seams and the surface pores can be detected by the LED images; the weld width and the fillet weld thickness can be measured by the structured light images.An automatic system based on machine vision is developed to inspect the quality of the weld seams on the horizontal arm. Machine vision technology is brought into the field of weld quality inspection, and high efficiency, high precision and intelligent automatic systems are developed. The optimal optical system is designed for obtaining clear original images (structured light images and LED images). The image processing algorithms are designed and optimized by the features of the structured light images and the LED images. The result shows that the system can detect and measure the main features of the weld seams efficiently and accurately. So it can be introduced to the weld quality inspection area,as a result, we can bring about real-time monitoring and management during the production process. |
PDF Full Text Request