Adaptive Image Recognition for Mixed Media Reality
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
There is a gap between printed media and electronic media, with no mechanism for publishers to allow access to electronic content using printed versions, and image recognition from low-quality images is difficult due to low-quality images produced by mobile devices and the computational expense of recognition processes.
Innovation Solution
A Mixed Media Reality (MMR) system that includes mobile devices, an MMR gateway, an MMR matching unit, and an MMR publisher, which uses recognition units and index tables to identify images and adapt to image queries, generating blurry images for training and modifying classifiers for improved recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If image recognition is performed on low-quality images from mobile devices, then accessibility to electronic content is improved, but recognition accuracy deteriorates
Solution Approach 1:
The system performs preliminary actions by generating multiple quality versions of reference images (original, blurred, low-resolution) and pre-computing recognition features for each version. When a query image is received, the system selects the appropriate reference quality level that matches the query image quality, thereby improving recognition accuracy for low-quality images without sacrificing the ability to handle high-quality images.
Solution Approach 2:
The system changes the quality parameter of reference images to match the query image. It dynamically selects between different reference image qualities (original, blurred, low-resolution) based on the detected quality of the input image. This parameter adaptation allows the recognition system to maintain high accuracy across varying image qualities from different mobile devices.
2Measurement precision
If comprehensive image recognition processing is applied, then recognition accuracy is improved, but processing time increases
Solution Approach 1:
The system segments the image recognition process into multiple quality levels and processing paths. Instead of applying comprehensive processing to all images, it divides the workflow into: (1) quick quality assessment, (2) selective reference image selection based on quality, and (3) targeted recognition processing. This segmentation allows the system to achieve high accuracy only when necessary, reducing overall processing time.
Solution Approach 2:
The system applies partial processing by performing only the necessary recognition steps based on image quality. For high-quality images, it uses full processing; for low-quality images, it selectively applies specific processing paths that are sufficient for that quality level. This avoids excessive processing and reduces unnecessary computational time while maintaining adequate accuracy.
3Reliability
If multiple recognition algorithms are used, then recognition robustness is improved, but system complexity increases
Solution Approach 1:
The system dynamically selects which recognition algorithms and reference image qualities to use based on the input image characteristics. Rather than statically deploying multiple algorithms for all cases, it adaptively chooses the appropriate algorithm combination and reference quality level for each specific query image, reducing complexity while maintaining robustness.
Solution Approach 2:
The system performs self-service by automatically assessing the quality of incoming images and autonomously selecting the appropriate reference images and processing paths. This self-adaptation eliminates the need for manual configuration of multiple algorithms for different scenarios, reducing system complexity while maintaining recognition robustness across varying conditions.
Data Source
AI summary
A MMR system for newspaper publishing includes a plurality of mobile devices, an MMR gateway, an MMR matching unit and an MMR publisher. The MMR matching unit receives an image query from the MMR gateway and sends it to one or more of the recognition units to identify a result including a document, the page and the location on the page. The image registration unit includes an indexing unit for generating images adapted to the environment and capabilities of the image capture device. The indexing unit also automatically adapts the configuration of the plurality of recognition units and index tables based upon image queries applied to the plurality of recognition and index tables. The plurality of recognition units and index tables are configured based on content they reference, recognition algorithm used or other factors.