Double-Angle Gradient Orientation for Thin-Line Preservation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining dominant gradient orientations in images fail to accurately preserve directional information while avoiding cancellation of opposing gradients, leading to loss of gradient information for thin-line structures and reduced angular resolution.
Innovation Solution
The method involves converting gradient samples into double-angle gradient vectors, combining them to form a compound gradient vector, and determining the dominant gradient orientation by converting this vector into a single-angle domain, preserving directional information and maintaining gradient orientation accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If gradient samples are averaged to determine local gradient vector, then directional information is obtained, but opposing gradients cancel out causing loss of gradient information for thin-line structures
Solution Approach 1:
The patent transforms gradient samples from single-angle domain to double-angle domain by doubling the angular component. This dimensional transformation allows opposing gradients (which differ by 180 degrees in single-angle domain) to be represented as identical vectors in double-angle domain, preventing cancellation during averaging while preserving directional information through the periodic nature of the transformation.
Solution Approach 2:
The patent changes the parameter representation of gradient vectors by transforming from single-angle parameters to double-angle parameters. This parameter transformation modifies how gradient directions are encoded, allowing the system to maintain both magnitude and directional information while avoiding the cancellation problem that occurs with conventional averaging of single-angle gradient vectors.
2Loss of information
If magnitude of gradient samples is taken before averaging, then opposing gradients cancellation is avoided, but directional information is lost
Solution Approach 1:
Instead of taking magnitude before averaging (which loses direction), the patent transforms gradient samples to double-angle domain where directional information is encoded in a way that survives averaging. The double-angle transformation preserves directional characteristics while enabling reliable averaging, as the transformed vectors maintain their directional relationships even when combined.
3Measurement precision
If absolute gradients are measured in multiple separate directions, then some direction information is recovered, but angular resolution is reduced to the number of sampled directions
Solution Approach 1:
The patent changes the parameter domain from discrete multi-directional sampling to continuous double-angle representation. By transforming all gradient samples into the double-angle domain, the system achieves high angular resolution without needing to sample in multiple discrete directions, as the transformation preserves continuous angular information in a form suitable for averaging.
Data Source
AI summary
Methods and image processing systems are provided for determining a dominant gradient orientation for a target region within an image. A plurality of gradient samples are determined for the target region, wherein each of the gradient samples represents a variation in pixel values within the target region. The gradient samples are converted into double-angle gradient vectors, and the double-angle gradient vectors are combined so as to determine a dominant gradient orientation for the target region.