Multispectral Imaging Element With On-Chip Spectral Signal Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Multispectral imaging elements require a large number of filters to achieve fine spectral characteristics, leading to a decrease in pixel resolution and increased processing time, which can hinder real-time processing in electronic devices like smartphones due to increased data amounts and complex calculations.
Innovation Solution
An imaging element configured as a semiconductor chip with an acquisition unit, pixel unit, conversion unit, processing unit, and output unit, featuring N types of pixels with different spectral characteristics, and capable of generating N+1 processed signals using information about the optical member and environment, allowing for improved spectral characteristics without the need for extensive filter arrays.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a large number of filters are provided to achieve fine spectral characteristics, then spectral resolution is improved, but pixel resolution decreases and the number of pixels per wavelength band decreases
Solution Approach 1:
The patent transitions from spatial arrangement of multiple filters to spectral processing in the signal domain. Instead of physically arranging many filters to achieve spectral resolution, the invention uses mathematical transformation (Fourier transform) on signals from a smaller number of filters to generate high-resolution spectral information, effectively moving the resolution enhancement from the spatial dimension to the frequency/spectral dimension.
Solution Approach 2:
The patent creates multiple spectral channels by computationally generating spectral information from a limited number of physical filters. Rather than requiring one physical filter per spectral channel, the system copies and processes signals through mathematical transformations to synthesize additional spectral views, achieving fine spectral characteristics without proportionally increasing the physical filter count.
2Measurement precision
If a large number of filters are provided to achieve fine spectral characteristics, then spectral resolution is improved, but processing time increases and real-time processing is hindered
Solution Approach 1:
The patent performs spectral processing operations within the imaging element itself, converting optical signals to electrical signals and applying Fourier transforms before data leaves the sensor. This preliminary action at the source eliminates the need for time-consuming post-processing of large datasets, enabling real-time spectral analysis by resolving the spectral information before the data burden accumulates.
Solution Approach 2:
The patent extracts and processes only the essential spectral information at the sensor level using Fourier transform operations, separating the critical spectral features from the full raw data stream. This extraction approach at the source reduces the data volume that requires extensive processing, thereby reducing processing time while maintaining spectral resolution.
3Measurement precision
If a large number of filters are provided to achieve fine spectral characteristics, then spectral resolution is improved, but data amount increases and power consumption increases
Solution Approach 1:
The patent generates multiple spectral channels through computational copying and transformation of signals from fewer physical filters. Instead of capturing and storing separate data streams from numerous physical filters, the system creates spectral copies through Fourier transform operations, achieving fine spectral resolution with significantly reduced data generation and storage requirements.
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 configuration enhances spectral characteristics while maintaining or improving pixel resolution and reducing processing time by performing spectral processing within the imaging element, thus optimizing data efficiency and power consumption.
Implementation Method 1
a pixel unit including N (N is an integer) types of pixels having different spectral characteristics with respect to a wavelength of input light input via the optical member
Data Source
AI summary
According to the present disclosure, there is provided an imaging element configured as a semiconductor chip, the imaging element including: an acquisition unit that acquires information regarding an optical member, the optical member being provided outside the semiconductor chip; a pixel unit including N (N is an integer) types of pixels having different spectral characteristics with respect to a wavelength of input light input via the optical member; a conversion unit that converts an output signal of the pixel unit into a digital output signal; a processing unit that performs conversion processing into N+1 or more processed signals having different spectral characteristics on the basis of an output signal output from the conversion unit by using the information; and an output unit that outputs a signal based on the processed signal to an outside of the semiconductor chip.