Product Interface Recognition System Based On Digital Tube Recognition

Since the mid-nineties of last century,a large number of technological changes have emerged in China,bringing a good opportunity for Chinese enterprises to comprehensively enhance their comprehensive competitiveness,it has also brought opportunities to China's household appliance market.However,despite the reputation of “the world's No.1 manufacture”,the current situation of appliance industry in China is still not optimistic.It is not difficult to find that although there are many household appliance enterprises in our country,the scale is relatively small and the level of technology production is relatively backward with less automation technology in the production process.Take product testing as an example,the product testing of the domestic small appliances industry is mainly based on manual testing which wastes time and energy.Thus,it is very important to realize the automation of product testing and improve the level of technology production,which is of great significance to the development of small household appliances and modernization.After analyzed the display interface of the product,this paper proposed and implemented a product interface recognition system based on digital tube recognition.In order to detect the interface display function of product in production line with real time and accurately,instead of traditional manual testing,improve the production efficiency and competitiveness of enterprises.This system includes digital tube localization,digital character segmentation and recognition,indicator light detection and recognition system implementation.Firstly,the image processing algorithm is used to pre-treatment,includes gray processing,median filtering,binaryzation,morphological processing,etc.Using the proper pre-treatment method can effectively remove the noise and enhance the target area,so the pre-treatment determines the subsequent accuracy of the follow-up.This paper proposed a rapid method of area positioning after analyzed the positioning methods of digital characters.Locate the character area by looking for the character's circumscribed rectangle.This method can effectively locate the digital character area according to the characteristics of the digital tube segment,and lay the foundation for later recognition.And then,this paper proposed an identification method for digital segment coding and location judgment,and compared it with the recognition algorithm based on Support Vector Machine.Experiment proved that this method is effective in identification of digital characters,included the slightly tilted case,it ensured the accuracy while greatly reducing the recognition time and improving the real-time performance of the algorithm.Finally,this paper introduced the detection algorithm of the indicator light outside the digital tube area in the product interface,and carries out the final system design and implementation.The method proposed in this paper can detect the display function of the product interface in real-time and effectively in dark environment.Experiment proved that the average recognition time of this method can be controlled within 120 milliseconds,the accuracy can meet the actual demand,and can be applied to the actual production line to realize the product Interface display test automation and increase productivity.