A filtering method for compressed video and a filtering device for compressed video
A filtering algorithm and video technology, applied in the direction of digital video signal modification, image communication, electrical components, etc., can solve the problem of only using the local spatial domain correlation of video images
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0089] figure 1 (a) is the flow chart of the filtering algorithm of the embodiment of the present invention, figure 2 It is a schematic diagram of the realization of the filtering method proposed by the embodiment of the present invention. Combine below figure 1 (a), figure 2 A specific explanation is given.
[0090] In step S106, the input is the image block P of size p×q currently to be processed m (See figure 2 box) and the reconstructed image, the output is a similar block third-order tensor Z, Z ∈ R p×q×K . First use the image block P m , in the reconstructed image, find K-1 similar blocks similar to the current block according to the similarity criterion, where the similarity criterion has different implementations, such as the sum of the absolute values of pixel differences in different image blocks (SAD, Sum of Absolute Differences):
[0091] SAD=||P m -P k(k≠m) ||
[0092] For example, the sum of squared differences of pixels in different image bloc...
Embodiment 2
[0114] figure 1 (b) shows a flowchart of the filtering algorithm according to the embodiment of the present invention. Combine below figure 1 (b) A specific description is given.
[0115] In step S102, the input video reconstruction image f and pixel position PP are received. In step S104, find out the image block {P t |t=1,2,...T}, where T is the number of image blocks passing through PP, P t in (i t ,j t ) corresponds to PP. For each image block P t , and the processes of steps S106 to S112 are performed respectively.
[0116] In step S106, the input is the currently pending p t ×q t size image block P t (See figure 2 box) and the reconstructed image, the output is a similar block third-order tensor Z t ,Z t ∈R p×q×K. First, using the current block as a template, find K-1 similar blocks similar to the current block in the reconstructed image according to the similarity criterion, where the similarity criterion has different implementations, for example, th...
Embodiment 3
[0142] figure 1 (b) shows the flowchart of the filtering algorithm of the present invention. Combine below figure 1 (b) Give a specific description.
[0143] In step S102, the input reconstructed video image f and pixel position PP are received. In step S104, find out the image block {P t |t=1,2,...T}, where T is the number of image blocks passing through PP, P t in (i t ,j t ) corresponds to PP. For each image block P t , respectively perform the processing of steps S106 to S112.
[0144] In step S106, the input is p t ×q t Image blocks of size P t (See figure 2 The box) and the reconstructed image, the output is a similar block third-order tensor Z t ,Z t ∈R p×q×K . Firstly, using the current block as a template, find K-1 similar blocks similar to the current block in the reconstructed image according to the similarity criterion. and (SAD, Sum of Absolute Differences):
[0145] SAD=||P t -P k(k≠t) ||
[0146] For example, the sum of squares of pixel d...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More