Automatic Recognition System Of Key Tooth Based On DSP

Lock assembly industry is the main component of national economy. Market capacity of locksales is huge and increases year by year. The present situation puts forward higher requirement tothe lock industry, with automation processing, the scale of production and processing efficiency.The traditional processing technology, manufacturing equipment and production mode have beenunable to meet the needs of the market. In recent years, research on the algorithm of imageprocessing technology changes rapidly, the process of image sensor is mature, the DSP processorperformance is powerful. Machine vision measurement system, based on DSP technology anddigital image processing and recognition, is widely applied to the transformation and upgrading ofmanufacturing technology.Key tooth recognition is a crucial procedure of core lock assembly, directly affects theefficiency and conformity rate. The current common methods are manual identification andsensor-mechanical identification. Manual identification has low efficiency and poor reliability,which is unable to meet the industrial production requirement. Sensor-mechanical identification hasdefects of mechanical loss, low recognition accuracy and automation degree. This research subjectcomes from Science and Technology Planning Project of Zhejiang Province. This paperinnovatively presents the solution of automatic recognition system of key tooth based on machinevision measurement technology. Using the characteristic of recognition accuracy and efficiency,non-contact, reliability and automaticity, focus on solving the problems with traditional recognitionmethods of low precision, poor automation degree, low machining efficiency, lock quality byhuman factors. The research has important significance in social practice. The recognition systemuses TMS320DM642as the core processor, image sensor MT9M001to acquire key image, theoperation of image processing algorithm to produce the key tooth code, CAN-bus to communicatewith the lock core assembly system. The image processing algorithm is developed based onOpenCV, then transplants to DM642platform. The main control system selects the correspondingpin tumbler according to the key tooth code, and then assembles together the pin tumbler, the lockcore with the lock body.The project studies and analyzes the development of DSP platform with TMS320DM642coreand design of key image processing algorithm based on OpenCV. Design the system schemeaccording to the technical requirement of the project. Introduce the design of system hardwarecircuit, and provide circuit schematic diagram. Secondly, design the key image processing and toothcode recognition algorithm. Include the image pre-processing, image rotation and coordinate transformation, tooth location, measurement and recognition. Under the CCS developmentenvironment, design the system software and the algorithm of transplantation and optimization.Finally, debug and test the system. Analyze the image recognition algorithm, statistical data andprocessing, performance analysis.The test result shows that the system works stable, the recognition accuracy and processingspeed can meet the technical design index of the subject.