Computer Vision Pre-processing via Structural Similarity Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing computer vision systems face challenges in consistently achieving successful image processing due to variations in image quality, as merely adjusting brightness or contrast does not guarantee successful processing by the computer vision processor.
Innovation Solution
A digital image processing method that computes the structural complexity of input images and compares it to true positive images stored in a database, updating control variables for signal quality conditioning processes to enhance the likelihood of successful computer vision tasks by adjusting brightness, noise reduction, gamma, and sharpness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If basic image conditioning operations (brightness, contrast adjustment) are performed on input images, then image visibility is improved, but computer vision task success rate does not consistently improve
Solution Approach 1:
The patent applies parameter changes by adjusting multiple image conditioning parameters (brightness, contrast, sharpness, noise reduction, gamma) based on the structural complexity characteristics of the input image. The system computes structural complexity metrics and uses them to dynamically determine optimal parameter values, transforming the image to match characteristics of known successful images rather than applying fixed or simple adjustments.
Solution Approach 2:
The patent performs preliminary action by pre-processing images through multiple conditioning operations before they enter the main computer vision processing pipeline. By computing structural complexity and applying appropriate transformations in advance, the system prepares images in a state that is more likely to succeed in subsequent vision tasks, effectively performing the conditioning work beforehand rather than relying on post-hoc adjustments.
2Measurement precision
If multiple image conditioning operations are applied to improve image quality, then processing accuracy improves, but system complexity increases
Solution Approach 1:
The patent manages complexity by systematically varying multiple parameters (brightness, contrast, sharpness, noise reduction, gamma) based on a unified structural complexity metric. Rather than independently tuning each parameter, the system uses the computed structural characteristics to drive coordinated changes across all parameters, reducing the effective complexity of the control problem while maintaining multiple degrees of freedom for optimization.
Solution Approach 2:
The patent implements feedback by using the computed structural complexity metrics as a basis for determining conditioning parameters. The system continuously assesses the input image's structural characteristics and adjusts the conditioning operations accordingly, creating a closed-loop approach where the image properties themselves guide the transformation process, thereby managing complexity through data-driven decision-making.
3Measurement precision
If structural complexity computation and image matching are performed, then pre-processing accuracy improves, but processing time increases
Solution Approach 1:
The patent extracts and utilizes only the essential structural complexity characteristics from the input image to drive the conditioning process. By focusing on key structural metrics rather than analyzing every pixel or feature, the system obtains sufficient information for accurate pre-processing while avoiding the computational burden of comprehensive image analysis, thus reducing processing time while maintaining accuracy.
Solution Approach 2:
The patent performs preliminary computation of structural complexity metrics before the main image conditioning operations. By calculating these guiding parameters in advance and using them to determine the appropriate conditioning strategy, the system avoids iterative or trial-and-error approaches during the actual processing phase, thereby reducing overall processing time while ensuring accurate parameter selection.
Data Source
AI summary
A processor computes a measure of input image structural complexity of an input image, and searches a database of true positives to find one or more entries in the database that represent true positive images that are structurally similar to the input image. The processor compares a measure of signal quality of the input image and a measure of signal quality of one of the true positive images, as retrieved from the database, and based on the comparison updates a control variable that configures a signal quality conditioning process that is to be performed on the input image prior to processing of the input image by a computer vision processor thus improving performance of the computer vision task. Other embodiments are also described and claimed.