As a project for the TMS320C80 we have implemented a video capture program. We have developed three methods and each of them has its special characteristics. The advantages and disadvantages of each method are described below.
In the first method we have the following process: first we initialize 'C80 and load the corresponding COFF file for this method and the video display task starts the execution. From this point we are able to pass a command to 'C80 to start the capturing process and 'C80 allocates all the available memory to its DRAM and fills it with the current displayed frames. Then we transfer the captured frames from SDB's DRAM to PC's RAM and we save it on the disk.
In the second method we have the following process: first we do the same initialization as in the first method and then we pass a command to 'C80 to capture and store a simple frame on its DRAM. After this we transfer this frame to PC's RAM and we save it on the disk. This process is repeated for the next frame.
The last method is the same with the second except that we don't save immediately the current captured frame but we keep it on PC's RAM until we have all the wanted frames.
Also we have tried to use the power of TMS320C80's parallel processors by implementing a 3D median filter on the video signal before we display it on the screen. The base process is described below:
Finally we have implemented a program which runs on the PC and works as an interface for passing commands to 'C80. Below you can see the two dialog boxes of this program.
Main dialog box
Display Adjust dialog box