Depth Map Generation from Single Image via Vanishing Lines
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for obtaining a depth map from a digital image are inefficient, often requiring multiple images, operator intervention, and result in long processing times.
Innovation Solution
A method that identifies horizontal or vertical planes in a single digital image, subdivides the image into homogeneous regions, and assigns depth levels based on chromatic characteristics, allowing for automatic generation of a depth map by merging partial maps according to image categories.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple images are analyzed to obtain depth information, then measurement precision is improved, but productivity deteriorates due to long processing times
Solution Approach 1:
The patent segments the depth map generation process into multiple passes: first generating a preliminary depth map using geometric constraints and vanishing lines, then selectively refining only uncertain regions through edge detection and local analysis. This segmentation allows rapid initial processing while maintaining precision where needed.
Solution Approach 2:
The patent applies partial action by performing comprehensive geometric analysis only on regions where it provides value, and using simpler methods elsewhere. The refinement stage selectively processes only regions with low confidence depth values, avoiding unnecessary computation in already accurate regions.
2Measurement precision
If operator intervention is used to identify objects in the image, then measurement precision is improved, but ease of operation deteriorates
Solution Approach 1:
The patent implements self-service through automatic vanishing line detection algorithms that identify geometric constraints without human input. The system automatically detects horizon lines, building edges, and perspective cues, then uses these to generate depth maps autonomously, eliminating the need for operator intervention while maintaining precision.
Solution Approach 2:
The patent replaces the mechanical process of manual object identification with automated computer vision algorithms. Edge detection, vanishing line detection, and depth calculation are performed by algorithms rather than human operators, substituting mechanical human cognition with automated computational processes.
3Productivity
If a single image is processed, then productivity is improved through faster processing, but measurement precision deteriorates due to limited information
Solution Approach 1:
The patent changes parameters locally within the image by detecting regions with different geometric characteristics. In regions with strong geometric cues (buildings, roads), it applies perspective-based depth estimation with high precision. In regions without such cues, it uses alternative methods like edge-based or gradient-based estimation, adapting the approach to local image characteristics.
Solution Approach 2:
The patent performs preliminary geometric analysis to identify vanishing lines and perspective structures before generating the depth map. This preliminary action extracts maximum depth information from geometric constraints, allowing accurate depth estimation from a single image without requiring multiple views or time-consuming sequential processing.
Data Source
AI summary
A method of obtaining a depth map from a single digital image of a three-dimensional scene containing geometric objects is provided that includes the identification in the image of at least one horizontal plane or one vertical plane of the three-dimensional scene and the assignment to every point of the image forming part of the identified plane, or to each one of the identified planes, of a depth level according to its distance from, respectively, a predetermined horizontal or vertical reference straight line of the image. In particular, the identification of at least one plane includes the identification in the image of at least two vanishing lines corresponding to two parallel lines in the plane and the assignment of a depth level made in accordance with the angular coefficients of the identified vanishing lines. Also included is a completely automatic method for obtaining a depth map from a single digital image of any three-dimensional scene.