Neural Network Feature Extraction for Image Encoding Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing demand for high-resolution and high-quality images leads to a significant increase in the amount of transmitted and stored information, resulting in higher transmission and storage costs, necessitating the development of high-efficient image compression technologies.
Innovation Solution
An image encoding/decoding method and apparatus utilizing an artificial neural network-based feature extraction method to improve encoding/decoding efficiency by signaling feature information, allowing for efficient transmission and reconstruction of bitstreams.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-resolution and high-quality image data are transmitted, then image quality is improved, but transmission cost and storage cost increase
Solution Approach 1:
The patent extracts and transmits only the most important visual information by identifying salient regions and salient objects within the image. Instead of transmitting the entire high-resolution image, the system extracts key features and regions that contain the most visually significant information, thereby reducing the amount of data to be transmitted while preserving essential image quality.
Solution Approach 2:
The patent applies different transmission qualities to different regions of the image based on their importance. Salient regions and salient objects are transmitted with higher quality and more detailed information, while less important regions are transmitted with lower quality or compressed more aggressively. This local differentiation allows the system to optimize the balance between image quality and transmission cost.
2Measurement precision
If the amount of transmitted information increases, then image quality is improved, but transmission cost increases
Solution Approach 1:
The system extracts only the essential visual information from the image by identifying salient regions and objects. By transmitting only this extracted essential information rather than the complete high-resolution image, the system significantly reduces the amount of data transmission required, thereby lowering transmission costs while maintaining the essential image quality needed for effective communication.
3Manufacturing precision
If feature information is signaled in detail, then encoding precision is improved, but device complexity increases
Solution Approach 1:
The patent segments the image into multiple regions based on their visual importance, identifying salient regions and non-salient regions separately. This segmentation allows the encoding system to apply different levels of detail and complexity to different parts of the image. By dividing the image into meaningful segments, the system can achieve high encoding precision for important regions while using simpler encoding methods for less important regions, thereby reducing overall device complexity.
Data Source
AI summary
An image encoding/decoding method and apparatus are provided. An image decoding method comprises obtaining, from a bitstream, encoded data of feature information generated by applying an artificial neural network-based feature extraction method to an image, reconstructing feature information by decoding the encoded data of the feature information, and generating analysis data of the image based on the feature information.