3D Model Construction from 2D Assets Using Contrast Enhancement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Photogrammetry systems face challenges in generating accurate 3D models from 2D images, particularly with highly reflective objects, as reflections can confuse point detection algorithms and obscure surface features, leading to inaccurate models.
Innovation Solution
The method involves performing local contrast enhancement and spectral suppression on 2D images to emphasize surface features and differentiate them from reflections, thereby improving the signal-to-noise ratio and generating a more accurate 3D model by isolating and amplifying high-frequency signals associated with object features and suppressing specular reflections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If photogrammetry is performed directly on 2D images of highly reflective objects, then the process is simple and fast, but the accuracy of 3D model generation deteriorates due to reflections confusing point detection algorithms
Solution Approach 1:
The patent applies preliminary image processing actions (local contrast enhancement and spectral suppression) to the 2D images before performing photogrammetry. This preprocessing highlights surface features and suppresses reflections in advance, ensuring that point detection algorithms can accurately identify features without being confused by reflections, thereby improving 3D model generation accuracy.
Solution Approach 2:
The patent applies local contrast enhancement to specific regions of the image where surface features are located, rather than uniformly processing the entire image. This selectively enhances the contrast of surface features in problematic areas while leaving other regions unchanged, improving feature detectability without unnecessarily increasing overall processing complexity.
2Difficulty of detecting and measuring
If local contrast enhancement is applied to emphasize surface features, then the detectability of features improves, but the processing time and computational resources increase
Solution Approach 1:
The patent applies local contrast enhancement selectively to specific regions of the image where surface features are located, rather than uniformly processing the entire image. This targeted approach improves feature detectability in critical areas while minimizing the overall computational burden and processing time.
Solution Approach 2:
The patent applies contrast enhancement with a degree that is sufficient to highlight surface features above the noise level created by reflections, without excessive enhancement that would unnecessarily increase processing complexity. The enhancement is calibrated to achieve the minimum necessary improvement in feature detectability.
3Measurement precision
If spectral suppression is used to remove reflection frequencies, then the signal-to-noise ratio improves, but the complexity of frequency analysis increases
Solution Approach 1:
The patent changes the spectral parameters of the image data by suppressing specific frequency ranges that correspond to reflection signals. This frequency-domain filtering modifies the signal characteristics to enhance the signal-to-noise ratio, allowing photogrammetry algorithms to more easily distinguish surface features from reflections.
4Measurement precision
If high frequency signals are amplified to highlight surface features, then the feature detection accuracy improves, but the amplification of noise also increases
Solution Approach 1:
The patent applies high-frequency signal amplification selectively to regions where surface features are present, rather than amplifying frequencies across the entire image. This localized amplification enhances feature detection accuracy in relevant areas while minimizing the amplification of noise in regions without features.
Solution Approach 2:
The patent converts the potentially harmful effect of noise amplification into a benefit by first suppressing reflection-related frequencies (which act as noise) and then amplifying high-frequency signals. The spectral suppression step removes harmful reflections, and the subsequent high-frequency amplification enhances surface features without proportionally amplifying the remaining noise.
Data Source
AI summary
Features of the surface of an object of interest captured in a two-dimensional (2D) image are identified and marked for use in point matching to align multiple 2D images and generating a point cloud representative of the surface of the object in a photogrammetry process. The features which represent actual surface features of the object may have their local contrast enhanced to facilitate their identification. Reflections on the surface of the object are suppressed by correlating such reflections with, e.g., light sources, not associated with the object of interest so that during photogrammetry, such reflections can be ignored, resulting in the creation of a 3D model that is an accurate representation of the object of interest. Prior to local contrast enhancement and the suppression of reflection information, identification and isolation of the object of interest can be improved through one or more filtering processes.