Image Processing Apparatus Data Hiding Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing techniques for image processing, such as adaptive reference sample smoothing (ARSS), face issues with reversible coding and subjective image quality when data hiding is applied, leading to reduced encoding efficiency and image quality, especially when transformation processes are skipped or residual DPCM is used.
Innovation Solution
An image processing apparatus and method that skips data hiding in specific conditions, such as during transformation or residual DPCM processes, to maintain image quality by controlling the hiding and decoding of predetermined data, including information for adaptive reference sample smoothing, transformation, and prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If data hiding is applied to improve prediction accuracy, then encoding efficiency is improved, but reversible coding is not guaranteed and lossless coding cannot be performed
Solution Approach 1:
The patent changes the parameter of data hiding application by introducing conditional logic: data hiding is applied only when transformation is performed, and skipped when transformation is skipped. This parameter change resolves the contradiction by ensuring reversible coding is maintained in lossless scenarios while preserving encoding efficiency gains in conventional scenarios.
Solution Approach 2:
The patent makes the data hiding process dynamic by adjusting its application based on the transformation flag state. When transformation is enabled, data hiding is applied; when transformation is skipped, data hiding is skipped. This dynamic adaptation allows the system to maintain both encoding efficiency and reversible coding guarantees under different operating conditions.
2Productivity
If data hiding is applied to improve prediction accuracy, then encoding efficiency is improved, but subjective image quality of decoded image is reduced
Solution Approach 1:
The patent changes the application parameter of data hiding based on the transformation flag. By conditionally applying data hiding only when transformation is performed, the patent prevents the degradation of subjective image quality that occurs when data hiding is applied to spatial domain data, while still maintaining encoding efficiency improvements in transformed domains.
Solution Approach 2:
The patent applies data hiding selectively to specific data domains (frequency domain after transformation) while excluding other domains (spatial domain when transformation is skipped). This local application approach ensures that data hiding benefits are obtained where appropriate without harming image quality in scenarios where transformation is not performed.
3Loss of information
If data hiding is applied when transformation process is skipped, then data in spatial domain is corrected, but subjective image quality of decoded image is reduced
Solution Approach 1:
The patent makes the data hiding process dynamic by linking it to the transformation flag state. When the transformation flag indicates transformation is skipped, data hiding is automatically skipped as well. This dynamic coupling prevents the harmful correction of spatial domain data while maintaining the ability to correct transformed domain data when appropriate.
Solution Approach 2:
The patent changes the operational parameter of data hiding based on the transformation process state. By monitoring the transformation flag and adjusting data hiding application accordingly, the patent eliminates the parameter combination that causes image quality degradation (data hiding on spatial domain data) while preserving beneficial parameter combinations.
4Measurement precision
If data hiding is applied, then prediction efficiency of RDPCM is reduced, but encoding efficiency is reduced
Solution Approach 1:
The patent changes the application parameter of data hiding based on the interaction between transformation and RDPCM processes. By conditionally applying data hiding only when transformation is performed (and implicitly when RDPCM benefits can be maintained), the patent resolves the contradiction by preventing the degradation of RDPCM prediction efficiency while preserving encoding efficiency gains in appropriate scenarios.
Data Source
AI summary
The present disclosure relates to image processing apparatus and method that can suppress a reduction of subjective image quality. A process related to hiding of predetermined data with respect to data regarding an image is executed, and hiding is skipped in a case where data in a spatial domain of the image is to be encoded. The data regarding the image for which the hiding is performed or the data regarding the image for which the hiding is skipped is encoded. The present disclosure can be applied to, for example, an image processing apparatus, an image encoding apparatus, an image decoding apparatus, and the like.