Baby steps

After lot of debates with my self i have to start my research project. The project is semantic video processing, analysis of the video data for search queries and automatic cataloguing it.

My hunch for taking it as topic for research is based on the following assumption

• Processor power has kept on increasing
  
• Network bandwidth has kept on increasing
  
• The compression technology is almost saturated
  

Now the ability to index the multimedia data will be in demand .

The tools i have chosen for my development is c and python. C because i can know it and python for rapid prototyping. I was also thinking of other option like perl, scilab, lush and all of them failed to meet my requirements. I not mentioned the most favourite or rather most talked about tool "Matlab ", it is bloated, it is slow and not open source. I always liked to work with simple tools which can be integrated together to preform complex activities.

Following are my requirements

• Must be fast
  
• Must be able to handle large data set
  
• must support image display ( i thought of writing the display unit but i realized i was recoding what was already available )
• Should have rudimentary support of numerical operation
  

I have started of with PSNR code in C. Since my development platform is linux i have to learn how to build and debug .

Now i have to take the second step , to bring in more focus and discipline in my work.