Video to Still Image Switching Encoding Apparatus
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Hardware still-image encoders face challenges in efficiently encoding still-image regions when switching from video to still-image regions, leading to increased information generation due to the inability to perform inter-screen prediction, resulting in higher data amounts compared to software encoders.
Innovation Solution
The proposed solution involves an encoding apparatus with a processor that performs intra-screen prediction encoding on a still-image region cut out from a decoded image before switching, and then uses this processed image for inter-screen prediction encoding after the switch, reducing the amount of generated information by referring to the previously decoded image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If a hardware still-image encoder is used to achieve higher-speed processing, then encoding speed is improved, but the ability to perform inter-screen prediction is lost resulting in increased information generation
Solution Approach 1:
The patent performs intra-screen prediction encoding in advance on the still-image region before switching from video to still-image mode. This preliminary encoding creates a decoded image that can be used as a reference for subsequent inter-screen prediction, allowing the hardware encoder to maintain encoding speed while enabling efficient compression in the first frame after switching.
Solution Approach 2:
The patent introduces an intermediary decoded image generated by intra-screen prediction encoding. This intermediary image serves as a bridge between the video mode and still-image mode, allowing the hardware encoder to reference this intermediate result for inter-screen prediction without requiring full software processing capabilities.
2Loss of information
If intra-screen prediction encoding is performed on still-image region before switching, then a reference decoded image is generated, but additional processing steps are required
Solution Approach 1:
The patent merges the intra-screen prediction encoding step with the existing encoding workflow for still-image regions. By combining this preliminary encoding with the subsequent inter-screen prediction using the generated decoded image, the system achieves efficient compression without requiring separate, complex processing pipelines.
3Loss of information
If inter-screen prediction is performed after switching from video to still-image region, then information generation is reduced, but the first frame after switching cannot utilize this method without reference images
Solution Approach 1:
The patent performs the necessary intra-screen prediction encoding as a preliminary action before the switching occurs. This ensures that when the first frame after switching arrives, the reference decoded image is already available, allowing immediate use of inter-screen prediction without delay or additional processing time.
Data Source
AI summary
An encoding apparatus, includes a memory; and a processor coupled to the memory and configured to: perform intra-screen prediction encoding on an image of a region of a still image cut out from a first decoded image corresponding to a screen image before a predetermined region is switched from a video to a still image, generate a second decoded image obtained by decoding information for which the intra-screen prediction is performed, and when a first screen image after the predetermined region is switched from a video to a still image is received, perform inter-screen prediction encoding on an image of a region of a still image cut out from the first screen image after switching to a still image, by referring to the generated second decoded image.