Region-of-Interest Video Compression for Facial Recognition Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image processing systems for remote cameras face challenges in accurate facial recognition due to image distortions caused by compression and decompression processes, leading to incorrect identification of non-face elements as faces, and require significant computing resources for pre-analysis at remote devices.
Innovation Solution
An apparatus for remote processing of raw image data that detects regions of interest, such as faces, and compresses them using a first compression level to maintain detail, while compressing non-interest areas differently or discarding them, facilitating accurate facial recognition at a central server without the need for extensive remote device computing capabilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If full image analysis is performed at the remote device before compression, then facial recognition accuracy is improved, but device complexity and computing cost increase significantly
Solution Approach 1:
The patent segments the image into multiple regions and classifies them as either regions of interest (containing faces) or non-interest areas. Different compression algorithms are then applied to different segments: high-quality compression for ROI and standard compression for non-ROI areas. This segmentation approach allows the remote device to perform only necessary analysis on critical regions rather than the entire image, reducing overall computational complexity while maintaining facial recognition accuracy.
Solution Approach 2:
The patent applies local quality by treating different parts of the image differently. Regions containing faces receive enhanced processing and higher quality compression to preserve facial features, while non-interest areas use standard compression. This localized approach ensures that computational resources are concentrated where they are most needed for accurate facial recognition, avoiding the need for high computing capability across the entire remote device.
2Productivity
If standard compression is applied to the entire image, then transmission efficiency is improved, but image quality and facial recognition accuracy deteriorate due to distortions
Solution Approach 1:
The patent divides the image into regions of interest and non-interest areas, then applies different compression algorithms to each segment. This segmentation allows the system to maintain high transmission efficiency for non-ROI areas while preserving image quality in critical facial regions, thereby maintaining facial recognition accuracy without sacrificing overall transmission efficiency.
Solution Approach 2:
The patent implements local quality by applying high-quality compression algorithms specifically to regions containing faces while using standard compression for the rest of the image. This ensures that facial features remain clear and recognizable for accurate facial recognition, while still achieving efficient compression ratios through the application of standard compression to non-interest areas.
3Quantity of substance
If high compression is applied to reduce data size, then transmission bandwidth requirements are reduced, but image distortion increases causing false face detections
Solution Approach 1:
The patent segments the image into regions of interest and non-interest areas, applying different compression levels to each. High compression is applied to non-ROI areas to minimize data size, while ROI areas receive reduced compression to preserve facial features and avoid distortion-induced false detections. This segmentation resolves the contradiction by allowing aggressive compression where it does not affect recognition accuracy.
Solution Approach 2:
The patent applies local quality by differentiating compression intensity across different image regions. Critical facial regions maintain higher image quality with less compression to prevent distortion and false face detections, while non-interest areas undergo high compression to minimize overall data size. This localized approach ensures reliable facial recognition without unnecessary computational overhead.
Data Source
AI summary
An apparatus for remote processing of raw image data receives the raw image data from a camera, such as a security camera. The apparatus includes a detection module to detect portions of the image data that contain possible regions of interest. Information indicating the portions that contain the possible regions of interest is then used during a compression process so that the portions that contain the possible regions of interest are compressed using one or more compression algorithms to facilitate further analysis and the remainder are treated differently. The compressed image data is then transmitted to a central system for decompression and further analysis. In some cases, the detection system may detect possible regions of interest which appear to be faces, but without performing full facial recognition. These parts of the image data are then compressed in such a way as to maintain as much facial detail as possible, so as to facilitate the facial recognition when it is carried out at the central server. The detection may be performed on the raw image data or may be performed as part of the compression process after a transformation of the raw image data has been carried out.