This project is a digital video codec system. By using a variety of different algorithms, the gains and losses of various trade-offs (various important parameters) are analyzed. Such as throughput, power consumption, cost, programmability, time to market, and application-specific aspects such as quality, target bit rate, latency, and error resilience.
Figure 1: Original_frame
These images are visual examples of video encoding analysis, demonstrating various encoding features and decisions. The left image shows the selection of a P-frame or I-frame with variable block size decisions overlaid. The middle image depicts the support for multiple reference frames, with each block marked by a color or other indicator showing the index of the reference frame used. The right image displays a map of motion vectors (MVs) overlaid on an arbitrary P-frame, using innovative representations to visualize the direction and magnitude of the motion vectors. These images help to deeply understand and analyze key technologies in the video encoding process.
After decoding, it will generate a reconsturcted YUV file. This figure shows 6 frames of Y component from YUV file. Then it will transform to RBG format.