Image Processing Clip Processing for Dynamic Range Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing methods face increased load in encoding and decoding due to dynamic range exceeding 16 bits during secondary or inverse secondary transformations, requiring larger intermediate buffers and increasing computational complexity.
Innovation Solution
An image processing apparatus and method that utilize clip processing to limit transformation coefficients within predetermined upper and lower bounds, and perform matrix computations and scaling to manage dynamic range, thereby reducing the load on encoding and decoding processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If secondary transformation is applied with 9-bit precision, then energy compaction is improved, but dynamic range exceeds 16 bits requiring larger buffers
Solution Approach 1:
The patent applies clip processing to transformation coefficients before they are used in subsequent processing stages. By preliminarily limiting the dynamic range of transformation coefficients through clipping, the system prevents the need for larger buffers while maintaining the energy compaction benefits of secondary transformation.
Solution Approach 2:
The patent changes the parameter of transformation coefficient precision by applying clip processing that limits values to a specific range. This parameter change allows the system to maintain 9-bit precision for energy compaction while preventing dynamic range from exceeding 16 bits, thus avoiding the need for larger buffers.
2Quantity of substance
If clip processing is applied to transformation coefficients, then dynamic range is controlled, but encoding and decoding load increases
Solution Approach 1:
The patent applies clip processing only to transformation coefficients that exceed the predetermined range, rather than processing all coefficients uniformly. This partial action approach controls dynamic range when necessary while minimizing the overall encoding and decoding load.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An image processing apparatus comprising a clip processing section configured to use one or both a predetermined upper limit and a predetermined lower limit to clip transformation coefficients obtained by applying a transformation process to a predicted residual that is a difference between an image and a predicted image of the image, wherein the predetermined upper limit and the predetermined lower limit are each decided based on a bit depth of an input signal including the transformation coefficients, wherein the predetermined upper limit and the predetermined lower limit are each further decided based on an extended computation precision flag of the input signal, and wherein the clip processing section is implemented via at least one processor.