IronDi51010
Newbie level 3
I am working on a senior design project that deals with image/video processing. Basic idea is to take a video of a person doing a series of hand gestures which are decoded to confirmed that they were performed. This is going to be performed on the altera DE2 with the Cyclone II FPGA. The image/video will be saved on SDRAM that is available on the board.
The problem is I do not know where to begin or what to look up with this project. I took DPS courses but the professor who taught those courses just gave the class a custom GUI he wrote that performs a series of fourier transform and signal manipulations. However, he never explained how the GUI worked, how to perform those transformation, or the math behind them. Just "Here is the expression to perform DFT. But I am not going to go over that as my GUI performs it." I tried to supplement my schooling by going through the DPS videos on MIT OCW. So I am familiar with the math and ideas like z-transform, FIR/IIR filters, etc.
I looked up information on google but everything I find is about MATLAB or general filters implemented in MATLAB.
The other crux is that my school only gives us 13 weeks for my major to design/build/test our projects. I have already spent 3 weeks just becoming familiar with VHDL and embedded C to write in NIOS. My goal is to decode video in real time, but I would like to process images in case I run short on time. That way I have something tangible to demonstrate.
A professor suggested to look at the first few pixels of the image but he refused to discuss it any further. What I am thinking is to save the pictures to SDRAM as RAW data and compare the captured image to a "finished image." Compare the captured image with the finished image for a specific sequence (e.g. "FF AB 33 ...") which would confirm that the specific sequence was taken. I plan on using the following camera daughter card: https://www.terasic.com.tw/cgi-bin/page/archive.pl?Language=English&CategoryNo=68&No=281 for time constraints.
I don't want to sound like I want an exact answer but what I am asking for guidance. Areas to look into, papers, tutorials. Something I can read to start going forward.
The problem is I do not know where to begin or what to look up with this project. I took DPS courses but the professor who taught those courses just gave the class a custom GUI he wrote that performs a series of fourier transform and signal manipulations. However, he never explained how the GUI worked, how to perform those transformation, or the math behind them. Just "Here is the expression to perform DFT. But I am not going to go over that as my GUI performs it." I tried to supplement my schooling by going through the DPS videos on MIT OCW. So I am familiar with the math and ideas like z-transform, FIR/IIR filters, etc.
I looked up information on google but everything I find is about MATLAB or general filters implemented in MATLAB.
The other crux is that my school only gives us 13 weeks for my major to design/build/test our projects. I have already spent 3 weeks just becoming familiar with VHDL and embedded C to write in NIOS. My goal is to decode video in real time, but I would like to process images in case I run short on time. That way I have something tangible to demonstrate.
A professor suggested to look at the first few pixels of the image but he refused to discuss it any further. What I am thinking is to save the pictures to SDRAM as RAW data and compare the captured image to a "finished image." Compare the captured image with the finished image for a specific sequence (e.g. "FF AB 33 ...") which would confirm that the specific sequence was taken. I plan on using the following camera daughter card: https://www.terasic.com.tw/cgi-bin/page/archive.pl?Language=English&CategoryNo=68&No=281 for time constraints.
I don't want to sound like I want an exact answer but what I am asking for guidance. Areas to look into, papers, tutorials. Something I can read to start going forward.
