Leaf Area Measurement Using DWA Path Planning and UNet Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for measuring leaf areas of dwarf plants are slow and lack accuracy.
Innovation Solution
A method utilizing a DWA path planning algorithm to determine a planned path for a robot, combined with a UNet model for leaf image segmentation and calculation, including an encoder, decoder, and trained segmentation network, to obtain accurate leaf areas.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional measurement methods are used for leaf area measurement, then the measurement process is simple, but the measurement speed is slow and accuracy is low
Solution Approach 1:
The patent replaces traditional mechanical measurement methods with an automated system combining mobile robots, RGB-D cameras, and deep learning algorithms (UNet model). The robot navigates to plants, captures images, and the UNet model performs automated segmentation and area calculation, eliminating manual measurement operations and significantly improving both speed and accuracy.
Solution Approach 2:
The patent changes the measurement parameters from manual visual estimation or contact measurement to digital image processing parameters. By capturing images with RGB-D cameras and processing them through the UNet model, the system extracts precise pixel-based measurements, transforming the measurement paradigm from analog to digital with higher precision and automation.
2Measurement precision
If deep learning technology is used for image segmentation, then segmentation speed and accuracy are improved, but system complexity increases
Solution Approach 1:
The patent employs the UNet model, a universal deep learning architecture originally designed for biomedical image segmentation, and adapts it for plant leaf segmentation. This pre-trained model can be fine-tuned for different plant types and conditions, providing a versatile solution that handles various segmentation challenges without requiring completely new system designs for each application.
Solution Approach 2:
The patent introduces an RGB-D camera as an intermediary device that captures both color and depth information. This intermediary provides rich feature data to the UNet model, improving segmentation accuracy while the model acts as another intermediary that processes this data and outputs precise leaf area measurements, distributing complexity across multiple specialized components.
3Productivity
If a robot is used for automated plant measurement, then measurement efficiency is improved, but path planning complexity increases
Solution Approach 1:
The patent implements preliminary path planning using the DWA algorithm before the robot executes measurement tasks. The robot pre-calculates optimal paths to multiple plants, avoiding obstacles and optimizing measurement sequences. This preliminary action ensures efficient navigation and minimizes idle time during actual measurements, boosting overall productivity without requiring complex real-time decision-making during image capture.
Data Source
AI summary
Provided is a method, system, and electronic device for calculating a leaf area of a plant, relating to the field of image processing. The method includes: obtaining a leaf image of a target plant on a planned path, where the planned path is a path determined based on a dynamic window approach (DWA) path planning algorithm; segmenting the leaf image by using a UNet model to obtain a leaf segmentation image, where the UNet model includes an encoder, a decoder, and a trained segmentation network that are connected to each other; and calculating a leaf area based on the leaf segmentation image. This application can achieve rapid and accurate segmentation of leaf images and calculation of leaf areas.