I-frames, p-frames and B-frames in MPEG images

MPEG image encoding consists of 3 components: I-frames, p-frames, and B-frames. During MPEG encoding, some images are compressed into I-frames, some are compressed into P-frames, others are compressed into B-frames. I-frame compression can get the compression ratio of the 6;1 without producing any perceptible ambiguity. I frame compression using P-frame compression, can achieve a higher compression ratio without perceptible blur phenomenon. B-Frame compression can achieve a compression ratio of 200:1, the file size is generally I-frame compression size of 15%, less than half of the P-frame compression size. I frame compression removes the spatial redundancy of the image, and the P-frame and B-frames remove the time redundancy, which is explained further below.

I frame compression uses the reference frame mode, only provides intra-frame compression, that is, the frame image is compressed to I-frame, only consider the image within the frame. I-Frame compression cannot remove inter-frame redundancy. Intra-frame compression is based on the discrete Cosine transform (DCT), similar to the compression standard used for DCT in JPEG and h.261 images.

The P-frame is predicted by using the predictive coding and the general statistic information of the neighboring frames. In other words, it considers motion characteristics and provides inter-frame encoding. The P-Frame predicts the difference between the current frame and the nearest I-frame or P-frame.
B-frames are bidirectional inter-frame encoding. It extracts data from the front and back I frames or P frames. B-frames are compressed based on the difference between the current frame and the previous frame and the next frame image.

MEPG a sample of uncompressed digital images of the SIF resolution specified by CCIR-601 at the beginning of the data stream. SIF resolution, for NTSC system, is the luminance signal is 352*240 each pixel, each chroma signal is 176*120 pixels. Each signal is 30 frames per second. The MPEG compressor determines whether the current frame is in an I-frame, p-frame, or B-frame. After the frame is determined, the DCT transform is used to quantify the results, rounding, and encoding the length of the stroke. The typical image frame sequence after encoding is: Ibbpbbpbbpbbibbpbbpbbpbbi ...

B-Frames and P-frames require that the computer be more powerful. Some compressors can not produce B-frame or even p-frames can not be generated, the image compression results will be very obvious discontinuity.