This thesis processes the work of developing CPU code and GPU code for Thomas Kaijsers algorithm for calculating the kantorovich distance and the performance between the two is compared. Initially there is a rundown of the algorithm which calculates the kantorovich distance between two images. Thereafter we go through the CPU implementation followed by GPGPU written in CUDA. Then the results are presented. Lastly, an analysis about the results and a discussion with possible improvements is presented for possible future applications.

<p>A transportation plan is a byproduct from the calculation of the Kantorovich distance between two images. It describes a transformation from one of the images to the other. This master thesis shows how transportation plans can be used for video coding and how to process the transportation plans to achieve a good bitrate/quality ratio. Various parameters are evaluated using an implemented transportation plan video coder.</p><p>The introduction of transform coding with DCT proves to be very useful, as it reduces the size of the resulting transportation plans. DCT coding roughly gives a 10-fold decrease in bitrate with maintained quality compared to the nontransformed transportation plan coding.</p><p>With the best settings for transportation plan coding, I was able to code a test sequence at about 5 times the bitrate for MPEG coding of the same sequence with similar quality.</p><p>As video coding using transportation plans is a very new concept, the thesis is ended with conclusions on the test results and suggestions for future research in this area.</p>

A transportation plan is a byproduct from the calculation of the Kantorovich distance between two images. It describes a transformation from one of the images to the other. This master thesis shows how transportation plans can be used for video coding and how to process the transportation plans to achieve a good bitrate/quality ratio. Various parameters are evaluated using an implemented transportation plan video coder. The introduction of transform coding with DCT proves to be very useful, as it reduces the size of the resulting transportation plans. DCT coding roughly gives a 10-fold decrease in bitrate with maintained quality compared to the nontransformed transportation plan coding. With the best settings for transportation plan coding, I was able to code a test sequence at about 5 times the bitrate for MPEG coding of the same sequence with similar quality. As video coding using transportation plans is a very new concept, the thesis is ended with conclusions on the test results and suggestions for future research in this area.