Design And Implementation Of LCD Symbol Recognition System In Automobile Instrument Detection

With the rapid development of automobile industry and electronic information technology,the automobile combination instrument is increasingly high degree of information technology and it's increasingly stringent quality requirements.Before putting into use,the automobile combination instrument need to be rigorously tested,but the efficiency and accuracy of manual testing is not suitable for automated production.In order to solve these problems,this paper designs an automobile combination instrument visual detecting system.After further research on the hardware design of the signal drive unit and the localization and recognition algorithm of the LCD symbol,this paper has been some research in the following areas.As a core hardware in the automobile instrument detection system,the signal driving unit is designed based on STM32 and it uses an integrated design pattern,The main controller controls the generation of all signals and the driving circuit is designed separately for each type of signal.After receiving the command from the host computer which the main controller has connected,the main controller can control the drive circuit to generate the signal.This integrated design pattern greatly improves the stability and precision of the system.For locating problem of instrument display area,this paper use an improved ACSS algorithm to roughly locate it,and then the edge detection operator is used to accurately locate the display area.The improved ACSS algorithm adds steps of filtering binary edges;the number of binary edges is reduced.After filtering the edge,a large number of false corner points are removed and the calculation is reduced.Compared with the current corner detection algorithm,the algorithm proposed in this paper is faster and more accurate.Based on locating the display area precisely,the BP neural network is used to recognize the LCD characters.Firstly,the binary image of the symbol is extracted after preprocessed,and then the projection method is used to segment the characters in vertical.Finally,the trained BP network is used to recognize the symbols.The results show that the recognition algorithm has high recognition rate and strong real-time performance.In addition,this paper also analyzes and realizes the recognition algorithms of other LCD symbol,LED and small table pointer angles.Finally,through the above work,this paper realizes the visual inspection system of automobile instrument,and the detection system can complete test cases.The system has strong expansion function and good compatibility,and has been applied in the automobile production line to achieve the expected effect.