Research and Development of Froth Flotation Video Surveillance Hardware Platform Based on DSP

Forth flotation is one of the most important methods of mineral separation. The appearance features of the foam is of vital importance for reflecting the production quotas of floating process, which has the role of guiding actual operation of mineral separation process. In present study, an embedded forth flotation video surveillance system is designed, which can not only reduce the labor intensity, subjectivity and the time consumed by monitoring caused by manual surveillance, but also solve the problems of low processing speed, high cost and difficulty in extension of the video surveillance based on PC.Regarding on background of mineral separation technology and combination of embedded system, we design a froth flotation video surveillance hardware platform based on TMS320DM642for faster processing speed, higher precision and smoother transmission of video information. The system can be divided into video capture module, the power clock module, processing module, memory module and the external interface module. An external video decoding circuit is developed in the video capture module to solve the problem of the inability of DSP in recognizing the HD analog signal, as well as to achieve the goal to convert analog signal to digital signal between HD analog camera and DSP rapidly. The monitoring of flotation process needs to collect large amounts of real-time foam images. But the on-chip SDRAM memory space is extremely limited. Thus we use the external high-capacity SDRAM to design the memory module, which realizes the seamless connection between SDRAM and DSP. In addition, SDRAM is a32-bit memory with slow storage speed. So two pieces of parallel SDRAM whose chip select signal are controlled by CEO are used to achieve high-speed64-bit data storage. Finally, based on the above hardware components of the monitoring system, the high-speed PCB is completed and the EMC testing is conducted.In order to ensure real-time communication between PC and the embedded system, the Ethernet interface with JPEG video signal compression method is used in the external interface module to solve the difficulty in transferring large amount of information, and to realize fast-Ethernet video transmission based on Socket programming. In terms of Bauxite forth flotation image, the watershed algorithm is applied to test the embedded foam video surveillance system. The experiment results prove that the system can collect and process HD foam video and features fast, at the meantime complete the transmission of lossless compression image to the host computer. The system meets the demand of the flotation monitoring process.45figures,5tables,5equations and60references are presented.