Robot Cleaner Boundary Detection Using Structured Light
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current robot cleaners face challenges in accurately detecting boundaries between objects, such as the floor and walls or obstacles, which are crucial for safe navigation, due to variations in lighting and object materials, leading to potential collisions.
Innovation Solution
A method and system for a robot cleaner that uses a light projection unit to project patterned light, an image acquisition unit to capture images, and a controller to apply a mask for detecting brightness differences between regions, allowing for the identification of seed pixels and feature detection to determine boundaries between objects, thereby controlling the robot's movement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional image analysis methods are used to detect boundaries, then the system is simple, but boundary detection accuracy deteriorates due to lighting variations and object material differences
Solution Approach 1:
The patent introduces a light projection unit as an intermediary device that projects structured light patterns onto the environment. This mediator actively illuminates boundaries between objects, creating high-contrast edges that are easily detectable by the image sensor, thereby significantly improving boundary detection accuracy without requiring complex image processing algorithms
Solution Approach 2:
The patent changes the illumination parameter by using active structured light projection instead of passive ambient light. By projecting specific light patterns and analyzing the reflected light, the system transforms the detection problem from one affected by varying ambient lighting conditions to one where the illumination is controlled and predictable, thereby improving detection accuracy
2Loss of time
If the robot cleaner uses basic image capture without active light projection, then the device complexity is low, but the detection time increases and collision prevention capability deteriorates
Solution Approach 1:
The patent applies preliminary action by projecting structured light patterns onto the environment before the robot makes movement decisions. This pre-illumination of boundaries allows the image sensor to capture high-contrast edge information in advance, enabling faster boundary detection and reducing the time required for safe navigation decisions
Solution Approach 2:
The patent replaces passive mechanical image capture with an active optical system that projects structured light. This substitution transforms the detection mechanism from relying on ambient light reflection to using controlled light projection and reflection analysis, significantly reducing detection time while maintaining relatively simple device architecture
3Reliability
If the robot cleaner does not use depth information from light patterns, then the system is simpler, but collision prevention capability deteriorates
Solution Approach 1:
The patent adds the depth dimension by analyzing the reflection patterns of projected structured light. By examining how the light patterns deform when reflected from surfaces at different distances and angles, the system extracts depth information that provides three-dimensional spatial awareness, significantly improving collision prevention capability
Solution Approach 2:
The patent implements feedback by continuously analyzing the reflected light patterns and using this information to adjust the robot's navigation in real-time. The system projects light, captures the reflection, processes the depth information, and uses this feedback loop to make immediate collision avoidance decisions, enhancing reliability
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
This approach enhances boundary detection accuracy, reduces detection time, and prevents collisions by specifying detection regions based on obstacles, while also providing more accurate depth information using the projected light pattern.
Implementation Method 1
a light projection unit (150) for downwardly projecting patterned light including a horizontal segment to an area in front of the main body (110)
Implementation Method 2
an image acquisition unit (120) for acquiring an image of the cleaning area
Data Source
AI summary
A robot cleaner and a control method thereof according to the present disclosure select a seed pixel in an image of an area in front of a main body, obtain a brightness difference between an upper region and a lower region obtained by dividing a predetermined detection area including each of neighboring pixels of the seed pixel and select a pixel belonging to a detection area having a largest brightness difference as a pixel constituting the boundary between objects indicated in the image. Accordingly, the boundary between objects present within a cleaning area can be detected rapidly and correctly through an image.