System Design Of Battery Plate Defects Detection Based On Line-scanning Camera

Along with the development of science and the improvement of the people’slivelihood, the batteries have entered all of our lives. In order to meet the sharpdemand for batteries, most of the battery plate production becomes wider andfaster in battery manufacturer today. Coating operation is needed in battery plateproduction, so the coated surface defect detection is an important link to ensure thequality of plates. However, most testing for coating defects still need artificial testto complete at present. Aiming at this problem, a new defect detection systembased on line-scanning camera is designed in the paper. This system can realizereal-time online detection and classification for battery plate coating defects.Firstly, the paper analyzes the working principle of system, and overall designdiagram is presented according to the functional requirements and technicalindicators. System can be divided into three parts: control circuit part, imageacquisition part and image processing and defect recognition part. Among them,after the two parts are running in the industrial control computer.Secondly, the hardware design and software programming of the controlcircuit are described in detailed. The control circuit which is designed usingTMS320F2812as main control chip contains speed measuring unit,communication unit and alarm unit.Thirdly, the paper introduces the design of image acquisition part whichadopts Visual Studio C++6.0as development environment. The setting of cameraparameters such as camera trigger mode, scanning frequency and the resolution ofthe adopt image is mainly completed in this part. Meanwhile, the image acquisitionand preservation thread, image pretreatment and exposure adjustment are alsoincluded in this part.Finally, the paper mainly introduces the design of the image processing anddefect recognition. Borland C++Builder6.0is used as development environment.Its main task is extracting images from the fixed path, processing image andjudging defect. It storages the defect parameters in order to record query in thefuture. In addition, this part designs the human-computer interaction interface forhuman operating.After half a year offline debugging and online operation, the system candetect the different defects effectively in the process of battery plate coating. Theoperation of the system has good reliability, stability, and high practical value aswell as can completely replace manual testing.