Dual-Light Camera Layout for Hyperspectral Fruit Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing camera systems for analyzing food products, particularly fruit-bearing plants, struggle with accurately capturing hyperspectral images due to variable natural light spectra and the inclusion of irrelevant light spectra, making it difficult to isolate desired portions for analysis and verification using AI models.
Innovation Solution
A camera system with two light sources, where the optical axes of the light sources intersect outside the camera's depth of field, combined with a line scan camera and robotic arm, allows for precise imaging and analysis of objects with a wide depth of field, using a halogen light source for hyperspectral imaging and RGB/depth cameras for localization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If natural light is used for hyperspectral imaging, then the imaging process is simple, but the variable spectra and irrelevant light spectra cause errors in processing and analyzing the image
Solution Approach 1:
The patent extracts only the relevant light spectra needed for fruit analysis by using a hyperspectral camera that captures specific wavelength ranges (400-1000nm), filtering out irrelevant light spectra from the natural illumination while maintaining the simplicity of using natural light
Solution Approach 2:
The patent changes the spectral parameters of the illumination by using halogen light sources that emit consistent spectra across the hyperspectral range, transforming the variable natural light into a more controlled illumination source while still maintaining relatively simple imaging conditions
2Productivity
If only hyperspectral images are captured, then the data processing is focused, but it is difficult to isolate desired portions of the fruit-bearing plant for verification and additional processing
Solution Approach 1:
The patent segments the imaging task into multiple components by capturing both hyperspectral images for detailed analysis and visible light images for contextual understanding and object isolation, allowing users to verify data by comparing different image types
Solution Approach 2:
The patent introduces visible light images as an intermediary that bridges the gap between hyperspectral data and human verification, making it easier to isolate and identify desired portions of the fruit-bearing plant by providing visual context
3Device complexity
If a single light source is used, then the device complexity is low, but the depth of field is limited and cannot accommodate objects at varying distances
Solution Approach 1:
The patent merges multiple light sources with different optical axes that intersect outside the depth of field, combining their illumination coverage to achieve extended depth of field while maintaining relatively simple device complexity
Solution Approach 2:
The patent extends the depth of field by arranging light sources in a spatial configuration where their optical axes intersect outside the depth of field, effectively adding a dimensional aspect to light source positioning that expands coverage without proportionally increasing complexity
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
Enables accurate and reliable hyperspectral imaging of fruit-bearing plants by isolating desired portions, allowing for effective data verification and additional processing through AI models, while minimizing light spectrum loss and ensuring precise object positioning.
Implementation Method 1
using a halogen light source for hyperspectral imaging
Implementation Method 2
a line scan camera...allows for precise imaging and analysis of objects
Data Source
AI summary
A camera system includes a camera, a first light source, and a second light source. A main optical axis of the first light source intersects with a main optical axis of the second light source at an intersection point, and the intersection point is not located within a depth of field of the camera.