Hyperspectral Image Reconstruction Using Prism Dispersion Model
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current hyperspectral imaging systems are expensive, bulky, and difficult to handle due to the need for specialized hardware like collimating optics and coded masks, limiting their usability and spatial resolution.
Innovation Solution
A method for reconstructing hyperspectral images using a conventional DSLR camera and a simple glass prism, which eliminates the need for collimating optics and coded apertures, and employs a novel image formation model and calibration method to estimate spatially-varying dispersion and reconstruct spectral information from sparse dispersion cues.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional hyperspectral imaging systems use collimating optics and coded masks, then spectral measurement capability is achieved, but device complexity and size increase significantly
Solution Approach 1:
The patent extracts and removes the collimating optics and coded masks from the traditional hyperspectral imaging system, retaining only the essential dispersive element (prism) and image sensor. This extraction eliminates unnecessary components while preserving the core spectral measurement functionality through direct projection of dispersed light onto the sensor.
Solution Approach 2:
The patent makes the prism serve multiple functions: it acts as both the dispersive element for spectral separation and the projection element for image formation. By eliminating dedicated collimating optics, the prism's function is extended to handle both spectral dispersion and spatial mapping, reducing overall system complexity.
2Measurement precision
If a large mask with many pinholes is used to isolate spectral dispersion, then spectral information is captured, but spatial resolution decreases significantly
Solution Approach 1:
The patent completely removes the large pinhole mask from the optical path, replacing it with a direct projection approach where dispersed light from the entire scene is mapped onto the image sensor. This eliminates the spatial resolution degradation caused by the mask while preserving spectral information through the prism's dispersion.
Solution Approach 2:
The patent transitions from a mask-based spatial filtering approach to a direct optical projection approach, changing the dimensionality of information capture. Instead of sampling through discrete pinholes (spatial dimension), the system captures continuous spectral information across the entire spatial field through the prism's angular dispersion mapped to spectral bands.
3Productivity
If multiple optical elements (prism, mirror, lens array) are used for snapshot image mapping, then spectral imaging is achieved, but device complexity increases
Solution Approach 1:
The patent merges the functions of multiple optical elements into a single prism. The prism simultaneously performs spectral dispersion and spatial projection that would otherwise require separate mirrors, lens arrays, and other optical components. This consolidation achieves snapshot spectral imaging with minimal elements.
Solution Approach 2:
The single prism in the patent performs multiple functions: it disperses light into spectral bands, projects the dispersed image onto the sensor, and enables snapshot capture without requiring separate optical paths for different spectral bands. This multi-functionality eliminates the need for mirrors, lens arrays, and other complex optical elements.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enables portable, cost-effective hyperspectral imaging with high spectral resolution and minimal impact on spatial resolution, allowing general users to capture hyperspectral information without advanced skills or complex setups.
Implementation Method 1
a dispersion model for dispersion created by a prism
Implementation Method 2
spectral dispersion by placing a large mask of pinholes in front of a prism
Data Source
AI summary
A method for reconstructing a hyperspectral image and a system therefor are provided. The method includes obtaining a dispersion model for dispersion created by a prism included in a camera, the prism including no coded aperture, and reconstructing a hyperspectral image corresponding to a captured image based on the dispersion model.