Image Detection via Down-sampling and Key Sub-region Feature Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current computer vision technologies face challenges in efficiently detecting lesion regions in medical images, such as diabetic retinopathy, due to the high computational requirements and time-consuming diagnostic processes, which can lead to delayed treatment.
Innovation Solution
A method involving down-sampling high-resolution images to obtain low-resolution images, determining key sub-regions, extracting features from both the low-resolution images and key sub-regions, and generating detection results based on these features to improve detection accuracy and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-resolution images are directly processed for detection, then detection accuracy is improved, but computational load and time consumption increase significantly
Solution Approach 1:
The image processing is segmented into two parts: down-sampled low-resolution images for global context and key high-resolution sub-regions for detailed feature extraction. This segmentation allows the system to process only critical areas at high resolution, improving detection speed while maintaining accuracy.
Solution Approach 2:
Different quality levels are applied to different regions: low-resolution processing for most areas and high-resolution processing only for key sub-regions containing potential lesions. This local quality approach reduces overall computational load while preserving detection accuracy where it matters most.
2Productivity
If high-resolution images are down-sampled to reduce computational load, then processing speed is improved, but information loss occurs
Solution Approach 1:
The system performs preliminary down-sampling to identify key sub-regions that may contain lesions, then retrieves and processes only those specific high-resolution areas. This preliminary action approach prevents information loss in critical regions while maintaining processing efficiency.
Solution Approach 2:
Key sub-regions containing potential lesion information are extracted from the down-sampled image and processed separately at high resolution. This extraction ensures that important diagnostic information is not lost during the overall down-sampling process.
3Measurement precision
If full high-resolution images are processed, then diagnostic accuracy is improved, but hardware resource consumption increases
Solution Approach 1:
Instead of processing the entire high-resolution image, the system processes only partial high-resolution key sub-regions that are most likely to contain diagnostic information. This partial action approach maintains diagnostic accuracy while significantly reducing hardware resource consumption.
Data Source
AI summary
Image detection methods and apparatuses, electronic devices and storage media include: down-sampling a second image to obtain a first image; determining at least one key sub-region in the second image for category detection; respectively extracting a feature of the first image and a feature of the at least one key sub-region; and generating a category detection result of the second image at least based on the extracted features.