Spectral Reflectance Correction Using 3D Depth and Light Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In short-distance optical measurements, structural changes and varying distances from the light source to different parts of the object cause inaccuracies in calculating spectral reflectance due to differing irradiance levels between the object and the diffuse reflecting plate, leading to errors in indirect measurement methods.
Innovation Solution
A method involving obtaining a registration depth map and spectral image, calculating three-dimensional coordinates and normal vectors, and using a full-light model to correct grayscale values based on depth information, ensuring equal irradiance across the diffuse reflecting plate and object for accurate reflectance calculation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If indirect measurement method using diffuse reflecting plate is used, then spectral reflectance can be calculated, but measurement precision deteriorates in short-distance scenarios due to irradiance differences between object and plate
Solution Approach 1:
The patent changes the parameter of irradiance distribution by using a light field model to calculate and compensate for irradiance differences at different spatial locations. Instead of assuming uniform irradiance, the system computes location-specific irradiance values based on depth information and light source geometry, thereby resolving the contradiction between using indirect measurement and maintaining accuracy in short-distance scenarios
Solution Approach 2:
The patent introduces a light field model as an intermediary between the diffuse reflecting plate measurement and the target object reflectance calculation. This model acts as a mediator that translates the plate's grayscale values into corrected reflectance values for the object by accounting for spatially varying irradiance, thus enabling accurate measurement without direct contact
2Measurement precision
If diffuse reflecting plate is placed close to object, then irradiance equality can be achieved, but device complexity increases due to positioning and pose alignment requirements
Solution Approach 1:
The patent replaces the mechanical positioning and alignment system with a computational light field model. Instead of physically adjusting the plate's position and orientation to match the object, the system uses depth maps and light source parameters to computationally determine irradiance distribution, eliminating complex mechanical adjustments while maintaining measurement accuracy
Solution Approach 2:
The patent transitions from two-dimensional plate positioning to three-dimensional spatial modeling by incorporating depth information from depth maps. This additional dimension allows the system to account for distance variations between the light source and different points on the object surface, resolving the contradiction without requiring precise mechanical alignment
3Ease of manufacture
If object surface is considered uniform plane, then calculation is simplified, but measurement precision deteriorates when surface structure varies
Solution Approach 1:
The patent segments the object surface into multiple discrete points with individual depth values from the depth map. Instead of treating the surface as a single uniform plane, the system divides it into element-wise depth points, each with its own irradiance calculation, thereby maintaining computational feasibility while accurately representing surface structure variations
Data Source
AI summary
A method includes: obtaining a registration depth map and a target spectral image; obtaining depth information of a target point from the registration depth map, and obtaining a first grayscale value of the target point; calculating 3D coordinate information of the target point according to the depth information of the target point, a first internal parameter of the imaging spectrum device, and first coordinate information of the target point in the target spectral image; determining normal vector information corresponding to the target point according to neighboring points of the target point; inputting the 3D coordinate information and the normal vector information into a full-light model of a diffuse reflecting plate to obtain a second grayscale value; and obtaining reflectance of the diffuse reflecting plate, and calculating reflectance of the target point according to the reflectance of the diffuse reflecting plate, the first grayscale value, and the second grayscale value.


