Image Processing Method Using Channel Expansion and Decomposition for Video Definition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In video transmission, especially during real-time audio or video calls, the compression of image information due to bandwidth limitations results in noisy and low-definition video images.
Innovation Solution
An image processing method using a convolutional neural network that performs channel expansion, multiple channel decomposition processes, and post-processing to enhance image definition, including feature extraction, downsampling, and dimension reduction, ultimately fusing intermediate images to improve video quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If image information is greatly compressed during video transmission to meet real-time requirements, then transmission bandwidth limitation is resolved, but video image definition deteriorates and noises increase
Solution Approach 1:
The patent segments the image processing task into multiple decomposition processes, dividing the input image into multiple decomposition images with different channel configurations. This segmentation allows the system to process and reconstruct image details more effectively, improving definition without requiring excessive transmission bandwidth.
Solution Approach 2:
The patent transforms the problem from spatial domain to channel dimension by performing channel expansion to obtain intermediate images with greater number of channels, then decomposing into multiple decomposition images. This dimensional transformation enables better feature extraction and image reconstruction, enhancing definition while maintaining transmission efficiency.
2Measurement precision
If channel expansion process is performed to obtain intermediate images with greater number of channels, then feature extraction capability is improved, but processing complexity increases
Solution Approach 1:
The patent decomposes the high-channel intermediate image into multiple decomposition images with different channel configurations. This segmentation reduces the complexity of processing each individual image while preserving the rich feature information obtained through channel expansion.
Solution Approach 2:
The patent performs channel expansion beyond the original channel count to obtain intermediate images with greater number of channels, then selectively decomposes them. This partial excessive action ensures sufficient feature extraction while the subsequent decomposition prevents overwhelming processing complexity.
3Manufacturing precision
If multiple channel decomposition processes are performed to extract high-dimensional features, then image definition enhancement is improved, but processing time increases
Solution Approach 1:
The patent performs channel expansion to obtain intermediate images with greater number of channels before decomposition. This preliminary action prepares the data in an optimal format that facilitates efficient subsequent decomposition processes, reducing overall processing time while maintaining definition enhancement.
Solution Approach 2:
By dividing the intermediate image into multiple decomposition images with different channel configurations, the patent enables parallel or sequential processing of smaller units, reducing the computational burden and processing time compared to processing a single high-dimensional image.
4Loss of information
If concatenated images are obtained by combining first and second decomposition images, then information retention is improved, but data volume increases
Solution Approach 1:
The patent concatenates first decomposition images from multiple processes with the second decomposition image from the last process to obtain a concatenated image. This merging combines complementary information from different decomposition stages, ensuring comprehensive information retention while organizing data in a structured manner for efficient processing.
Data Source
AI summary
An image processing method and device, and a computer-readable storage medium are disclosed. The method includes: performing a channel expansion process on the input image to obtain a first intermediate image; performing a channel decomposition process for multiple times based on the first intermediate image, wherein each time of channel decomposition process includes: decomposing an image to be processed into a first decomposition image and a second decomposition image; concatenating first decomposition images generated in each time of channel decomposition process and second decomposition image generated in the last time of channel decomposition process to obtain a concatenated image; performing a post-processing process on the concatenated image to obtain a second intermediate image; and fusing the second intermediate image with the input image to obtain the first output image.