Encoding Device Embedding ROI Information in Pixel Values
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing systems face challenges in synchronizing Region of Interest (ROI) information with decoded images, especially when data is lost in communication channels, leading to difficulties in matching the decoded image and ROI information in their correspondence relation.
Innovation Solution
An encoding device embeds ROI information into pixel values within the image, allowing the decoding device to read and synchronize the information without requiring simultaneous input, using an embedding position designation unit to specify where the ROI information is embedded, and a post-filtering unit to restore dynamic range.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If ROI information is transmitted separately from the decoded image through communication channels, then the system can provide region of interest control, but data loss in communication channels causes synchronization difficulties between the decoded image and ROI information
Solution Approach 1:
The patent merges ROI information with the image data by embedding it directly into the encoded bitstream. The ROI setting image is created by superimposing ROI information (such as coordinate data or mask images) onto the original image, and this combined ROI setting image is then encoded together with the main image data. This ensures that the ROI information and image data remain synchronized throughout transmission and decoding, eliminating synchronization issues caused by separate transmission channels.
Solution Approach 2:
The patent introduces an ROI setting image as an intermediary element that carries both image data and ROI information in a unified format. This ROI setting image serves as a mediator that integrates the functionality of separate image transmission and ROI annotation, allowing the decoding device to extract both the image and ROI information from a single synchronized source without requiring separate communication channels.
2Productivity
If an encoder with video quality control parameters for respective blocks is used, then ROI coding can be achieved efficiently, but the encoder cost and complexity increase
Solution Approach 1:
The patent performs preliminary processing by creating an ROI setting image that marks the region of interest before the actual encoding process. This ROI setting image contains pre-defined information about which regions should receive higher quality encoding. The encoder then uses this pre-prepared ROI setting image to guide the encoding process, allocating more bits to ROI regions without requiring complex real-time quality control mechanisms during encoding.
Solution Approach 2:
The patent changes the encoding parameters dynamically based on the ROI setting image. By analyzing the ROI information from the setting image, the encoder adjusts encoding parameters such as quantization step size, transform block size, and bit allocation for different regions of the image. This allows standard encoders to achieve ROI coding functionality by simply changing how they apply existing encoding parameters, rather than requiring fundamentally different encoder architecture.
3Quantity of substance
If preprocessing filtering is applied to non ROI regions to reduce information amount, then code amount is reduced, but image quality in non ROI regions deteriorates
Solution Approach 1:
The patent applies different encoding qualities to different regions of the image based on the ROI setting image. Non-ROI regions are encoded with lower quality parameters (such as larger quantization steps or coarser transforms) to reduce code amount, while ROI regions are encoded with higher quality parameters to maintain image quality. This local differentiation of encoding quality allows the system to optimize the overall code amount while preserving quality where it matters most.
Data Source
AI summary
An encoding device sets a region of interest corresponding to an input image, and prepares, as an intermediate image, data in which the input image is associated with region-of-interest information, designates an embedding position, embeds the region-of-interest information in a pixel value form into the intermediate image in accordance with the embedding position, prepares, as a region-of-interest setting image, data associated with the region-of-interest information, encodes the image by using a predetermined image coding scheme, and outputs an encoded image. A decoding device analyzes a decoded image obtained through decoding, reads the region-of-interest information, prepares, as a region-of-interest information read image, data in which the decoded image is associated with the region-of-interest information, applies a post-filter on the basis of the region-of-interest information read image, and outputs a post-filtered image.