Research Of Voice Diaphragm Image’s Parallel Processing Based On Lifting Wavelet Transform And Implementated On FPGA

With the development of the computer technology,digital image processing technology and microelectronics technology, machine vision detection technology has obtained fast development. Machine vision equipments plays a more and more important role in the manufacturing production field, and has been used more and more extensively.The traditional machine vision detected technology which bases on PC has certain advantages in the implement of digital image processing algorithm,but it limits the speed of voice diaphragm detection and reduces the production efficiency.Therefore, there is a new way which shows how to detect the voice diaphragm with a parallel image processing method based on lifting wavelet and realize on FPGA in this thesis. This method can accurately extract the inner circle and outer circle of voice diaphragm and can reach a high-speed of online real-time detection.Using this method,companies will reduce production cost and improve the qualified rate of the speaker.It also improves production efficiency,economic benefit and the competitiveness of the enterprise.In this paper, with the small voice diaphragm as the research object, lifting wavelet transform as the main theory and FPGA as the hardware chip for algorithm,the main research contents of this paper include the following respects:1. Introduce the research background and significance of this thesis and present research status of voice diaphragm detected technology and digital image parallel processing technology.Find out the advantages and disadvantages of the present voice diaphragm detected technology and the advantages of digital image parallel processing technology.2. Analyse the basic theory of wavelet transform knowledge,including development course of wavelet transform, continuous wavelet transform, discrete wavelet transform, multi-resolution analysis and Mallat algorithm.Study the basic theory, advantages and implemented steps of lifting wavelet transform.3.Give a algorithm of detecting the inner circle and outer circle of voice diaphragm.The principle and steps of this algorithm are introduced in detail. Validate the correctness of the algorithm by using Matlab software. Comparing with a few classic edge detected algorithm (Sobel, Canny, etc.),we can find out advantages of the algorithm of detecting the inner circle and outer circle of voice diaphragm.4. Realization of the concentricity detected algorithm on FPGA.Introduce the digital image processing technology based on FPGA and the software and hardware environment of the experiment. Then explain the principle, implement on FPGA and timing simulation of each module of the algorithm. Finally,the algorithm procedure is written into the FPGA development board SEED-XDTK-V4.In the experiment,the model R0810H of voice diaphragm is used to debug and validate the algorithm,then analyses the datas of the result and shows a good detective effectiveness.