Spectral Filter Mosaic Material Identification Without Demosaicing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for automatic vehicle occupancy detection using multi-band infrared cameras with mosaic spectral filters require demosaicing operations, which increase processing time and data storage, and are inefficient for identifying materials in images.
Innovation Solution
A method that bypasses demosaicing by determining material indices for each spectral filter cell using intensity values from neighboring sensor elements, allowing for material identification directly from patterned images captured by multi-band IR cameras with a mosaic of spectral filter cells arrayed in a geometric pattern.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If demosaicing operation is performed on raw image data from mosaic filter patterned sensor arrays, then complete spectral information for each pixel is obtained, but processing time and data storage requirements increase
Solution Approach 1:
The patent extracts only the necessary spectral information from the mosaic filter patterned image by determining material indices for each spectral filter cell using intensity values from neighboring sensor elements, without performing complete demosaicing. This extraction approach obtains sufficient spectral data for material identification while avoiding the time-consuming full demosaicing process.
Solution Approach 2:
The patent performs preliminary material identification by calculating material indices directly from the patterned image data before complete demosaicing would be required. By establishing material indices for each spectral filter cell using available neighboring intensity values, the system achieves material identification without waiting for full image reconstruction.
2Measurement precision
If demosaicing operation is performed on raw image data from mosaic filter patterned sensor arrays, then complete spectral information for each pixel is obtained, but data storage requirements increase
Solution Approach 1:
The patent extracts only the essential spectral characteristics needed for material identification by determining material indices for each spectral filter cell, rather than storing complete demosaiced images. This approach retains sufficient spectral information for accurate material identification while significantly reducing data storage requirements.
3Ease of manufacture
If single-band infrared imaging is used, then system cost is reduced, but robustness against object occlusion and posture variations deteriorates
Solution Approach 1:
The patent makes the single-band infrared imaging system perform multiple spectral analysis functions by using a mosaic filter patterned sensor array with multiple spectral filter cells. Each filter cell captures different spectral information, enabling multi-band material identification capabilities in a single-band system, thereby improving robustness against occlusion and posture variations without increasing system cost.
Solution Approach 2:
The patent adds spectral dimensionality to single-band infrared imaging by incorporating mosaic spectral filter cells that capture different wavelength information. This transforms a single-dimensional (intensity-only) imaging system into a multi-dimensional system that captures spectral signatures, enabling more reliable material identification under varying conditions.
4Device complexity
If visible light is used for occupancy detection, then system simplicity is maintained, but reliability under environmental conditions deteriorates
Solution Approach 1:
The patent substitutes visible light detection with infrared detection, replacing a simpler but less reliable system with a more sophisticated sensor type that operates in the infrared spectrum. This substitution enables detection through tinted windows and under various environmental conditions while maintaining automated detection capability.
Solution Approach 2:
The patent changes the operational wavelength parameter from visible light to infrared spectrum, allowing occupancy detection to function reliably under environmental conditions such as tinted windows, rain, snow, and darkness where visible light detection fails. The infrared wavelength parameter enables penetration through obstacles that block visible light.
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 reduces image processing time and data storage requirements while enabling accurate material identification in vehicle occupancy detection systems, improving robustness against environmental factors and object occlusions.
Implementation Method 1
Each spectral filter cell filters source light by a wavelength range of interest
Implementation Method 2
Each spectral filter element is aligned with a sensor element of the camera's sensor array to collect separate filtered intensity values for each pixel location in the image
Implementation Method 3
a multi-band infrared camera system with a mosaic of spectral filter cells arrayed in a geometric pattern
Data Source
AI summary
What is disclosed is a system and method for processing image data acquired using a multi-band infrared camera system with a spectral mosaic filter arranged in a geometric pattern without having to perform a demosaicing that is typical with processing data from an array of sensors. In one embodiment, image data that has been captured using a camera system that has a spectral filter mosaic comprising a plurality of spectral filters arrayed on a grid. A material index is determined, using intensity values collected by sensor elements associated with a cell's respective spectral filters. All of the material indices collectively generate a material index image. Thereafter, material identification is performed on the material index image using, for example, pixel classification. Because the demosaicing step can be effectively avoided, image processing time is reduced. The teachings hereof find their uses in a wide array of applications including automated HOV/HOT violation detection.