Monocular Multispectral Imaging With Single-Path Filter Registration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing multispectral sensors on unmanned aerial vehicles face challenges such as high cost, complex image synchronization, and registration precision issues, as well as inefficiencies in image data acquisition and processing.
Innovation Solution
A monocular high-resolution multispectral imaging system with a distortionless imaging lens, multi-waveband combined narrow-band optical filter, and parallel data processing unit, which includes an image sensor to convert optical signals into digital signals, and a self-stabilizing gimbal for precise targeting, to generate reflectivity images for vegetation analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a multi-lens sensor is used to capture multispectral images, then waveband selection flexibility is improved, but sensor cost increases significantly and image registration precision deteriorates
Solution Approach 1:
The patent combines multiple narrow-band optical filters (red, green, blue, yellow-green, red-edge, near-infrared) onto a single imaging lens, merging multiple spectral filtering functions into one integrated optical path. This eliminates the need for multiple separate camera lenses while maintaining waveband selection flexibility and reducing system complexity and cost.
Solution Approach 2:
The single imaging lens with multiple narrow-band filters serves multiple functions: it captures images across six different spectral wavebands simultaneously, providing both structural and spectral information through one optical path. This multi-functional design replaces what would traditionally require multiple specialized lenses.
2Measurement precision
If a beam-splitting sensor is used to separate light rays into several wavebands, then image capturing and registration are improved, but light intensity loss increases and signal-to-noise ratio deteriorates
Solution Approach 1:
Instead of using beam splitters to separate light into different wavebands, the patent extracts only the necessary spectral information by placing narrow-band optical filters directly in the optical path of a single lens. This approach captures full-resolution images for each waveband without the light loss associated with beam splitting, while maintaining precise image registration since all filters share the same optical path.
3Adaptability or versatility
If a filter wheel is used to capture multispectral images, then waveband flexibility is improved, but imaging speed decreases and mechanical wear increases
Solution Approach 1:
The patent replaces the mechanical filter wheel system with a static array of narrow-band optical filters positioned in the optical path. This eliminates moving mechanical parts entirely, allowing for faster imaging without wear and tear, while maintaining the ability to capture multiple wavebands simultaneously through the single lens.
4Volume of moving object
If a compact multispectral sensor with optical filters is used, then device size is reduced, but spectral sensitivity decreases and image processing difficulty increases
Solution Approach 1:
The patent applies narrow-band optical filters at specific locations in the optical path where they can selectively transmit desired spectral wavebands while blocking others. Each filter is positioned to optimize its spectral transmission characteristics, maintaining high spectral sensitivity for the target wavebands (red, green, blue, yellow-green, red-edge, near-infrared) while keeping the overall sensor compact.
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
The system achieves lower cost, improved image synchronization, and higher registration precision, enabling efficient crop monitoring with reduced data storage needs and flexible spectral band selection for various agricultural scenarios.
Implementation Method 1
a multi-waveband combined narrow-band optical filter, placed between the distortionless imaging lens and the image sensor and configured to constrain a spectral range of an actual receivable optical signal of the image sensor
Implementation Method 2
an image sensor, configured to convert an optical signal of each waveband into a digital signal
Implementation Method 3
a distortionless imaging lens, configured to focus an original image optical signal of a target object on the image sensor to generate a red-green-blue (RGB) original image
Data Source
AI summary
Disclosed are a monocular high-resolution multispectral imaging system and a control method, relating to the technical field of vegetation index monitoring. The imaging system includes: an image sensor, configured to convert an optical signal of each waveband into a digital signal; a distortionless imaging lens, configured to focus an original image optical signal of a target object on the image sensor to generate a red-green-blue (RGB) original image; a multi-waveband combined narrow-band optical filter, placed between the distortionless imaging lens and the image sensor and configured to constrain a spectral range of an actual receivable optical signal of the image sensor; an incident light calibration module, configured to acquire incident irradiance of ambient light; and a parallel data processing unit, connected to the image sensor and configured to analyze the digital signal in combination with the incident irradiance and generate a reflectivity image for vegetation analysis. It is conducive to solving the problems of high cost as well as insufficient image synchronization and registration precision of existing multispectral sensors.