Robot Mower Visual Control for Missed High Grass Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Visual self-moving devices, such as robot mowers, often use random mowing modes that result in missed grass areas, which are not efficiently processed, leading to taller growth and spread to surrounding areas.
Innovation Solution
A control method for self-moving devices that involves obtaining and processing images of the forward area to determine the presence of specific objects like high grass, allowing the device to move directly towards and process these areas, using binocular visual modules and image processing techniques like parallax processing and color segmentation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a random mowing mode is adopted, then the device can operate simply without complex path planning, but the traversal efficiency is low and grass is missed in some places
Solution Approach 1:
The patent implements feedback by using the visual module to continuously detect grass height in real-time during mowing operations. The controller receives visual feedback about missed grass areas and dynamically adjusts the mowing path to revisit and process these areas, transforming the simple random mode into an adaptive system that improves traversal efficiency while maintaining operational simplicity.
Solution Approach 2:
The patent applies preliminary action by using the visual module to scan and identify areas with high or missed grass before the mowing blade reaches them. This advance detection allows the controller to plan and adjust the mowing path proactively, ensuring that problematic areas are targeted efficiently rather than being discovered only after random passage.
2Reliability
If ultrasonic module is used to detect obstacles, then collision prevention is achieved, but high missed grass is misidentified as obstacles and left unprocessed
Solution Approach 1:
The patent applies local quality by assigning different detection functions to different modules: the ultrasonic module handles general obstacle detection for collision prevention, while the visual module specifically targets grass height identification. This localized functional differentiation allows the system to distinguish between true obstacles and missed grass areas, enabling selective processing without compromising collision safety.
Solution Approach 2:
The visual module serves as an intermediary between the ultrasonic obstacle detection system and the mowing execution system. It provides additional information about grass height and vegetation characteristics, allowing the controller to differentiate between actual obstacles that require avoidance and missed grass areas that require re-processing, thus resolving the misidentification problem.
3Device complexity
If missed grass is not processed timely, then the device maintains simple random mowing patterns, but missed grass grows taller and spreads to surrounding areas
Solution Approach 1:
The visual module provides continuous feedback about grass height and distribution patterns, enabling the controller to identify and track missed grass areas over time. This feedback mechanism allows the system to dynamically adjust mowing patterns to target persistent problem areas, preventing grass from growing taller and spreading while maintaining relatively simple operational complexity.
Data Source
AI summary
A control method for a self-moving device controls the self-moving device to move in a designated area to process a predetermined object in the designated area. The method includes steps of: obtaining, during the moving process of the self-moving device, an image of an area in a forward direction of the self-moving device; and determining, according to the image, whether the predetermined object in the area in the forward direction of the self-moving device includes a specific predetermined object, so as to control a moving direction of the self-moving device. The specific predetermined object and the predetermined object have at least one different feature parameter. A related self-moving device is also disclosed.