Semantic Segmentation for Differential Video Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video compression methods fail to efficiently reduce data size for video files used in vehicle interior monitoring without compromising the quality of relevant information, particularly in applications like car sharing and taxi services, where economically efficient transfer and storage are crucial.
Innovation Solution
A method utilizing learning-based semantic segmentation to differentiate between monitoring and periphery areas in video images, allowing for higher quality compression of relevant areas while reducing the data rate of less important areas, such as windows or vehicle exteriors, using a combination of neural networks and traditional image processing techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If uniform high-quality compression is applied to the entire video image, then the quality of relevant information is maintained, but the data size and storage costs increase
Solution Approach 1:
The patent applies different compression qualities to different regions of the video image based on their semantic importance. The learning-based segmentation model identifies monitoring areas (requiring high quality) versus periphery areas (tolerating lower quality), enabling localized quality adjustment that maintains relevant information quality while reducing overall data size
Solution Approach 2:
The video image is segmented into multiple regions with different compression requirements using a learning-based semantic segmentation method. This divides the image into monitoring areas and periphery areas, allowing differential compression strategies to be applied to each segment, thus resolving the contradiction between maintaining quality and reducing data size
2Quantity of substance
If compression is applied to reduce data size, then storage and transfer costs decrease, but the quality of essential information may be degraded
Solution Approach 1:
The system ensures essential information quality by applying high compression quality specifically to regions identified as containing relevant information (monitoring areas), while applying lower compression quality only to non-essential periphery areas. This localized quality control prevents degradation of essential information while achieving overall data size reduction
3Ease of manufacture
If all areas of the video are compressed with the same quality, then the processing is simple, but economically efficient compression is not achieved
Solution Approach 1:
The video processing pipeline is segmented into distinct stages: semantic segmentation to identify important regions, followed by differential compression applied to different regions. This structured segmentation maintains processing simplicity through clear separation of concerns while enabling economically efficient compression by reducing data size in non-critical areas
Data Source
AI summary
A method for generating a monitoring image. The method includes: providing an image sequence of the surroundings to be monitored with the aid of an imaging system; determining at least one monitoring area and at least one periphery area of at least one image of the image sequence with the aid of a learning-based semantic segmentation method; compressing the monitoring area of the at least one image of the image sequence with a first compression quality; and compressing the periphery area of the at least one image of the image sequence with a second compression quality to generate the compressed monitoring image, the second compression quality being lower than the first compression quality.