R&D Of Embedded Machine Vision System Based On STM32F407

Machine vision device in industrial production is applied more and more widely, requirements are also getting higher and higher, the machine vision device or PC based, high power, large volume, difficult installation, poor or the real-time performance of the algorithm and performance index, or the whole set of equipment is costly. In view of these problems, this paper develops an embedded machine vision device. The device adopts STM32F407 as the core processor with floating point operation, and the image acquisition device is CMOS OV7670 color image sensor, and the external RAM chip cache image data.. The paper studies the development of the embedded machine vision device and the application of the image positioning, and selects the soles images for the localization experiments, and obtains the better effect.The main research contents are as follows:1. Architecture design embedded machine vision deviceFirst, the working principle embedded machine vision device, then according to functional requirements and performance equipment designed architecture for embedded machine vision device and the functional modules.2. Hardware design of embedded machine vision deviceAccording to the architecture and function modules means required to determine specific hardware device selection and implementation, and the corresponding hardware circuit design, while briefly working principle of the system hardware of each module are described. Including the main processor module, image acquisition module, real-time display design and test modules, memory modules, SD card module and communication interface module circuit.3. Software embedded machine vision system designFirst introduced the software main program flow, and then the device functional requirements and STM32F407 and COMS OV7670 work, the software routine process equipment and its implementation are described in detail. It includes:initialization, image acquisition and pre-processing, storage and transmission of image data, the client’s remote settings, and other image processing algorithm image sensor.4. Debugging and results analysis embedded machine vision deviceFirstly, the image acquisition and display functions were tested, and then the client software to the test, and finally to the sole image positioning were tested. Through specific experiments, analyze test results, identify problems and propose methods for improvement. After extensive testing, we used to verify the design meets the functional requirements of device design and performance.Finally, the paper summarizes the full text, the main research results are explained, and points out the shortcomings of this paper and the improvement of the research.