Research On System Of Digital Image Proccessing And Recognition Of Instruments

With the development of times, the progress of the society, the constantly update of technology, traditional artificial meter reading is no longer suitable for the needs of society development and will be washed out gradually, instead, automatic meter reading system will be widely development. For a variety of automatic meter reading ways, this paper deeply studies digital image meter reading, designs a set of digital image acquisition, processing and recognition system, this system can be suitable to the development of intelligent building in our country.The work of this paper mainly includes the following three parts, the first part of digital image acquisition; The second part of digital image processing, The third part digital image recognition.Digital image acquisition:In order to get a clear digital image which can be used in processing and recognition, as master control chip, the MC9S12XS128freescale MCU controls OV6620image sensor to acquisite the digital image of instrument, and then the task of digital image acquisition is fulfilled, as a bridge between MCU and PC, communication com completes data communications between MCU and PC.Digital image processing:In order to achieve a effect which is most suitable for image recognition, and further enhance the recognition rate of digital recognition algorithm, this paper uses conservative smooth、mean filtering、gaussian sharpen and binarization four kind of digital image processing methods, and finally get a clear black-and-white digital image.Digital image recognition:In order to be able to find a simple digital image recognition algorithm which is the most suitable for this system, this paper uses edge tracking method, feature discriminant method, distance feature information method and feature information extraction method, four kinds of recognition algorithms to recognize digital image which was processed, select out the digital image recognition algorithm which is the most suitable for this system according to overall recognition rate which is achieved by experiment, so as to get the digital image recognition algorithm applied to this system.This system finish the work of digital image acquisition, processing and recognition, and ultimately achieve digital image recognition algorithm which is suitable to this system by a lot of experiment tests.