Floor-Cleaning Robot Vision Control for Targeted Dirt Detection

The patent introduces an imaging sensor as an intermediary device between the robot and the floor surface. This sensor captures images of the floor, which are then processed to identify areas with significant dirt or debris. The imaging sensor acts as a mediator that translates visual information into actionable navigation data, enabling the robot to detect dirt without requiring complex direct sensing mechanisms.

The patent replaces traditional mechanical or simple contact-based sensing methods with an optical imaging system. Instead of using mechanical sensors to detect dirt presence, the system uses cameras to capture images and processes these images computationally to identify dirty areas. This substitution of mechanical sensing with optical-electronic sensing enables more precise dirt detection while maintaining manageable system complexity.

The patent applies local quality by directing cleaning efforts selectively to specific areas of the floor that are identified as having significant dirt or debris. Instead of uniform cleaning across the entire floor area, the system processes images to locate dirty regions and generates navigation commands to target these specific locations. This localized approach to cleaning improves efficiency by concentrating resources where they are most needed.

The patent implements preliminary action by capturing images of the floor area before cleaning and processing these images to identify dirty regions. This pre-scan and pre-planning phase allows the robot to formulate a targeted cleaning path that prioritizes areas needing attention. By performing this preliminary detection and planning, the robot avoids wasting time on already clean areas and proceeds directly to problematic regions.