Embedded Monitoring System Of High Definition Froth Image

Image capture and analysis is the core module of monitoring system in flotation process, and most of the existing systems capture standard definition (SD) froth image for analysis and processing, which results in image feature can not be accurately extracted and seriously affect the optimization control of the flotation process. In this paper, the capture and processing of high definition (HD) foam image is studied, and analysis system based on DaVinci technology for HD foam image is designed. The main work of this paper is as follows:Firstly, aiming at the difficulties in extracting foam image feature due to low-resolution images is weak on performance the details of foam under harsh image capture environment, HD foam image monitoring system based on DaVinci technology, unique dual-core ARM+DSP architecture, is designed for HD foam image capture and process. System hardware, including the TMS320DM6446main chip modules, memory modules, HD image capture module, network transmission module, debug module, et al, is designed and described.Secondly, considering the problem of capture HD images in embedded environment, on the basis of analysis the current mainstream HD image capture interface technology, a HD image capture scheme based on the embedded GigE Vision technology is proposed, and related hardware circuit is designed, GigE Vision protocol is parsed. Then, software development technology for the TMS320DM6446dual-core processor is studied, including loading BootLoader, porting Linux operating system, configuration Codec Engine and other features that make the ARM and DSP TMS320DM6446able to work together.Finally, this paper adopts the GigE Vision protocol used for the ARM kernel processing combined with DMA to complete the collection of high-definition images, and the DaVinci shared memory technology is utilized to exchange high-definition image data with the DSP; in the bubble image feature extraction algorithm based on the DSP nuclear transplantation, the inline functions owned by DaVinci and software pipelining technology are employed to achieve the transplantation and optimization of froth image feature extraction algorithms.Testing results show that the system is able to effectively capture and extracting HD froth images with high processing accuracy and speed.38figures,7tables and62references are presented.