Dynamic Light Adjustment in Machine Vision for Plant Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing machine vision algorithms for agricultural applications struggle with varying light and weather conditions, leading to inaccurate plant detection and requiring cumbersome physical solutions or high-powered lighting systems.

Innovation Solution

A dynamic adjustment method that combines color threshold algorithms with neural networks, using asynchronous processing to intermittently calibrate color threshold parameters based on environmental changes, incorporating growth-stage-specific adjustments, reference data from calibration elements, and light sensors to maintain accurate plant detection.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If neural net based algorithms are used for plant detection, then detection accuracy may be improved, but computational resources and power consumption increase significantly

Engineering Contradiction:

Improveplant detection accuracyVSAvoidpower consumption

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The system dynamically adjusts between neural network and color threshold algorithms based on environmental conditions. Color threshold algorithms are used during normal operation for low power consumption, while neural networks are activated only when lighting conditions change or calibration is needed, optimizing the trade-off between accuracy and power usage

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The neural network executes periodically or intermittently to calibrate color threshold parameters rather than continuously. This periodic execution maintains detection accuracy through regular calibration while significantly reducing overall computational load and power consumption compared to continuous neural network operation

Inventive Principle:

Principle #19Periodic action

2Use of energy by moving object

If color threshold algorithms are used for plant detection, then power consumption is reduced, but detection accuracy decreases under varying light conditions

Engineering Contradiction:

Improvepower consumptionVSAvoidplant detection accuracy

Core Design Contradiction:

Use of energy by moving objectVSMeasurement precision

Solution Approach 1:

The system uses feedback from environmental light sensors and periodic neural network evaluation to dynamically adjust color threshold parameters. When light conditions change, the neural network evaluates current thresholds and provides feedback for recalibration, ensuring color threshold algorithms maintain accuracy across varying conditions while keeping power consumption low

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The system changes parameters of the color threshold algorithm based on environmental conditions. Color thresholds are dynamically adjusted according to lighting conditions, plant growth stage, and neural network evaluation results, allowing the simple algorithm to maintain high accuracy without increasing computational complexity

Inventive Principle:

Principle #35Parameter changes

3Reliability

If hoods or high-powered lighting systems are added to control lighting conditions, then detection reliability is improved, but device complexity and installation burden increase

Engineering Contradiction:

Improvedetection reliabilityVSAvoidinstallation complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The system replaces mechanical solutions (hoods, physical light barriers) with software-based adaptive algorithms. By using dynamic parameter adjustment and intelligent algorithm selection, the system achieves reliable detection under varying conditions without adding physical complexity to the imaging setup

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

Instead of physically controlling light with hoods or high-powered lights, the system changes algorithmic parameters to adapt to existing lighting conditions. This software-based adaptation maintains detection reliability while avoiding the complexity of additional hardware components

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20250308233A1Dynamic Light Adjustment for Machine Vision

Publication Date: 2025.10.02 AG LEADER TECHNOLOGY INC

AI summary

A method for dynamically adjusting machine vision for agricultural applications includes capturing, via at least one camera mounted on an agricultural vehicle, image data of plants in an agricultural field, obtaining reference data from a calibration element within the field of view of the at least one camera, wherein the calibration element provides a baseline for visual comparison under changing environmental lighting conditions, detecting a change in environmental lighting conditions by comparing current visual characteristics of the calibration element to predetermined baseline visual characteristics, calculating adjustment values for parameters of a machine vision algorithm based on the detected change in environmental lighting conditions, modifying the parameters of the machine vision algorithm according to the calculated adjustment values, and processing the image data with the modified machine vision algorithm to identify plants in the agricultural field.