Safe Area Monitoring with Modified Obstacle Boundaries

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional anti-collision systems in vehicles inaccurately determine obstacle boundaries, leading to excessive sensitivity and false warnings due to discrepancies between identified and actual obstacle areas, which can disrupt driving safety.

Innovation Solution

A monitoring method that generates a modified area close to the obstacle boundary using a pre-established image identification model, allowing for a two-stage determination of obstacle invasion by comparing the safe area with this modified area, reducing erroneous determinations and enhancing accuracy and reliability.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If object detection technology is used to identify obstacles, then the monitoring system can detect obstacles, but the identified obstacle area is always larger than the real area, leading to excessive sensitivity and false warnings

Engineering Contradiction:
Improveobstacle boundary identification accuracyVSAvoidfalse warning rate
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent introduces a modified safe area as an intermediary concept between the originally detected safe area and the obstacle boundary. This modified safe area is generated by shrinking the original safe area by a predetermined distance, creating a buffer zone that filters out false positives while maintaining sensitivity to real threats. The determination logic checks whether obstacle boundaries invade this modified safe area rather than the original safe area, thereby reducing false warnings caused by detection inaccuracies.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Area of stationary object

If the safe area is enlarged to improve detection coverage, then more obstacles can be detected, but the system becomes excessively sensitive and generates more false warnings

Engineering Contradiction:
Improvesafe area coverageVSAvoiddetermination accuracy
Core Design Contradiction:
Area of stationary objectVSReliability

Solution Approach 1:

The patent segments the safe area determination into two distinct stages: first, the original safe area is established for comprehensive coverage; second, a modified safe area is generated by shrinking the original safe area to create a more conservative monitoring zone. This segmentation allows the system to maintain large coverage area while using the shrunk modified safe area for actual invasion determination, thereby resolving the contradiction between coverage and accuracy.

Inventive Principle:
Principle #1Segmentation

3Productivity

If object detection technology is used, then obstacles can be identified, but the identified obstacle area exceeds the real area, causing unnecessary safety mechanisms to be triggered

Engineering Contradiction:
Improvemonitoring efficiencyVSAvoiddisturbance to driver
Core Design Contradiction:
ProductivityVSObject-generated harmful factors

Solution Approach 1:

The modified safe area serves as a mediator that filters out false positive detections. By requiring obstacle boundaries to invade this shrunk modified safe area rather than the original larger safe area, the system maintains efficient monitoring while significantly reducing false warnings that would otherwise disturb the driver or trigger unnecessary safety mechanisms.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP4648025A1Monitoring method for safe area and safety system and displaying method thereof
Publication Date: 2025.11.12 WHETRON ELECTRONICS
  • EP4648025A1 patent drawingFigure 1
  • EP4648025A1 patent drawingFigure 2~3
  • EP4648025A1 patent drawingFigure 4

AI summary

A monitoring method for a safe area, performed by a processing module (2) having a processor and a pre-established image identification model, includes the following steps. An input image (I) is generated corresponding to a predefined viewing direction of a driving target (T), and the input image (I) has an image corresponding to an obstacle (O). A safe area (SA) is generated in the input image (I). A predetermined route (R) is generated, and a continuous area that does not belong to the obstacle from a predetermined range extended from the predetermined route (R) is defined to generate a modified area (MA), so that in a situation that the predetermined range includes a part of an obstacle (O), a part of a boundary of the modified area (MA) corresponds to a part of a contour of the obstacle (O). Then a step is executed to determine whether the safe area (SA) is entirely included in the modified area (MA).