Image Encoding Quantization Parameter Adjustment for Bit Rate Stability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
During Constant Bit Rate (CBR) control in video encoding, region-based image quality control can cause a rapid increase in bit rate due to large deviations in generated code amounts, especially when the size of specified regions changes rapidly, such as during camera direction changes.
Innovation Solution
An image processing apparatus that sets a specified region with a quantization parameter and changes the quantization parameter to a predetermined fixed value when the region changes, ensuring the image quality is lower than before, to prevent a sudden increase in bit rate, while maintaining a target code amount.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If region-based image quality control is performed during CBR control, then image quality inside the specified region is improved, but bit rate may instantaneously become very large when the size of the specified region rapidly changes
Solution Approach 1:
The patent applies preliminary action by detecting changes in the specified region before encoding proceeds. When a change is detected, the quantization parameter for the entire image is preemptively adjusted to a value that prevents excessive bit rate increase, rather than waiting for the bit rate problem to occur. This proactive adjustment ensures smooth bit rate transitions while maintaining region-based image quality control.
Solution Approach 2:
The patent dynamically changes the quantization parameter based on the state of the specified region. When the specified region changes (expands, contracts, or moves), the quantization parameter is adjusted to a predetermined value that prevents bit rate from becoming excessively large. This parameter change strategy balances image quality requirements with bit rate constraints during CBR control.
2Stability of the object's composition
If the quantization parameter is adjusted to maintain target code amount, then bit rate stability is improved, but image quality may deteriorate when the specified region changes rapidly
Solution Approach 1:
The patent applies local quality by differentiating between regions during encoding. The specified region maintains higher image quality with a separate quantization parameter, while the outside region uses a different quantization parameter determined by CBR control. When the specified region changes, the outside region's quantization parameter is adjusted to maintain overall bit rate stability, creating a localized solution that preserves both image quality and bit rate control.
Solution Approach 2:
The patent uses feedback from the specified region's state to adjust encoding parameters. The detection unit monitors changes in the specified region, and based on this feedback, the quantization parameter is dynamically adjusted. This feedback mechanism ensures that when the specified region changes, the system responds by adjusting parameters to maintain bit rate stability while minimizing impact on image quality.
Data Source
AI summary
An image processing apparatus includes a setting unit configured to set, within an image, a specified region which is associated with a set quantization parameter, a determination unit configured to determine a quantization parameter to be used in encoding of an image of an outside of the specified region, in such a manner that a generated code amount obtained by encoding the image becomes a target code amount, and a changing unit configured, in a case in which the specified region is changed when the determination unit is activated, to change a quantization parameter of the image to a predetermined fixed value in such a manner that an image quality of the image becomes lower than a previous image quality obtained from a quantization parameter set for the specified region before the change.