The Contour Size Detection System Research Of Automotive Airbag Based On Image Recognition

In the research of a car’s safety performance, as an important component to keep the driver safe, the airbag’s quality has been paid much attention all the time. The key point of the research to keep the driver safe is whether the quality of the airbag is good or bad and if it can pop out in time when a traffic accident happens, so it needs to be done to check the loading size of the airbags and make it sure that the airbags are produced with high quality.Six sigma is a comprehensive and systematic management approach to improve production process quality, through specialized quality improvement methods to achieve the goal of quality improvement and process, to eliminate the unstable factors in the process of production the impact on the process output. Six sigma management is aprevention management, its statistical analysis by collecting data of manufacturing information, identify potential risks in production, and develop a positive prevention programs to prevent bad product production, in the process of production to ensure product quality, reduce the output eventually rejected product. And defined as the key beginning of six sigma to improve the quality of production process, whether its work is done well influences the whole process improvement workdirectly, andthen for the next step of measurement and analysis process with the preparation.Because the airbag has high requirements for production line and it should use the production technical such as squeeze, press and so on, and the airbag should be covered with plastic material, so its contour dimension is the important part to be checked in the product quality test. So far, most factories mainly use mechanical dial gauge to take some outline border points to test the positive or negative deviation, but this manual inspection can only check a part of the products and the time to do the test is relatively long, and the accuracy is not high. Besides, as this kind of artificial detection equipment can’t do the real-time detection in the assembly line of the airbag, it not only wastes time, but also can’t improve productivity. In order to adapt to the new flexible automatic assembly line, and to meet the more strict demands of quality control, this paper designs a kind of detecting system for automobile safety airbag outline size based on image recognition. This detection system uses the principle of machine vision detection technology, and uses the special image processing module of industrial computer with high performance, selects various of logic operation and edge detection methods to complete the image of the airbag, through the correlation values to determines whether the product is qualified in the range of error.Experiments show that the detection time of the new detection system is very short, it shortens testing station’s time a lot and improves production efficiency. And its detection accuracy is very high, and it can fully meet the requirement of the production quality, moreover, it improves the sampling type positioning error detection into a full inspection type image identification and detection, and can completely meet the requirements of six sigma quick definition, so it is very suitable for the airbag assembly production line with real-time online detection.