Self-Walking Mower Image Segmentation for Boundary Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing intelligent mower systems require buried electrified boundaries, which are costly and limit the shape of lawns, making it necessary to develop a more convenient self-working system that can be built on the ground.
Innovation Solution
A control method for a self-walking device that processes captured images by acquiring pixel values, segmenting the image into sub-regions, calculating their sizes, and determining if the image belongs to a lawn or non-complete lawn region based on preset thresholds, allowing the device to navigate and avoid obstacles effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a buried electrified boundary is used to define the working region, then the intelligent mower can sense the boundary and navigate automatically, but the cost of manpower and material resources increases and the shape of the lawn is limited
Solution Approach 1:
The patent replaces the mechanical/buried electrified boundary system with an optical/image processing-based boundary detection system. The intelligent mower uses its camera to capture images of the ground, and the processor analyzes these images to identify boundary lines marked on the ground surface. This substitution eliminates the need for buried electrified wires and complex installation procedures while maintaining reliable boundary sensing capability.
2Ease of operation
If a buried electrified boundary is used to define the working region, then the intelligent mower can sense it and navigate automatically, but the cost of installation increases
Solution Approach 1:
The patent extracts the boundary detection function from the buried electrified boundary system and implements it through image processing. Instead of relying on physical buried wires, the system extracts visual information about boundaries from camera images and processes this information to achieve automatic navigation. This extraction approach eliminates the need for extensive material resources while preserving the automatic navigation capability.
3Ease of manufacture
If image processing is used to detect boundaries on the ground, then the installation cost and complexity decrease, but the accuracy of boundary detection must be maintained
Solution Approach 1:
The patent applies preliminary image processing operations including noise reduction, edge detection, and line fitting algorithms before making the final boundary determination. By performing these preliminary actions on the captured images, the system enhances the precision of boundary detection even though the boundary is simply marked on the ground rather than being a physical barrier. This preliminary processing ensures accurate boundary identification despite the simplified installation method.
Data Source
AI summary
A self-working system, a self-walking device (1) and a method for controlling same, and a computer-readable storage medium. The control method comprises: acquiring a captured image; processing the captured image to acquire a processed image; segmenting the processed image into at least one sub-region; calculating the size An of each sub-region, respectively; counting the number of sub-regions with An>V in the processed image, and marking same as the number Nb of special sub-regions, wherein V is a preset quantity threshold; if Nb≤1, judging that the captured image belongs to a lawn region; and if Nb> 1, judging that the captured image belongs to a non-complete lawn region. If it is judged that a captured image belongs to a non-complete lawn region, it can be determined that there is a large obstacle or a boundary (2), etc. Whether the self-walking device (1) encounters an obstacle or a boundary (2) can be analyzed by analyzing a captured image, such that the operation is easier, and the control is more sensitive and effective.