Vehicular Acceleration Suppression Using Bird's-Eye View Parking Frame Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing vehicular acceleration suppression systems inaccurately activate throttle suppression when a vehicle enters a parking space, degrading drive performance by mistakenly interpreting normal driving actions as erroneous accelerator manipulations.

Innovation Solution

A vehicular acceleration suppression device that captures images of the surrounding environment, converts them to a bird's-eye view, and extracts parking frame candidates to accurately determine when to suppress acceleration based on the driver's intended actions within the parking frame, eliminating candidates that do not meet predefined length thresholds.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If throttle suppression is activated when the vehicle departs from the road based on map information, then erroneous accelerator manipulation is prevented, but drive performance in parking spaces degrades

Engineering Contradiction:
Improveaccuracy of erroneous manipulation detectionVSAvoiddrive performance in parking space
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent introduces an image recognition system as an intermediary to verify whether the vehicle is actually in a parking space before suppressing throttle. The system captures images of road markings, detects parking frame lines, and uses this visual information to confirm the vehicle's location context, thereby preventing false suppression during legitimate parking maneuvers while maintaining protection against erroneous acceleration on open roads

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent segments the road departure detection into two independent verification paths: (1) map information-based detection and (2) image recognition-based detection. By dividing the detection system, each method can specialize - map data provides broad contextual awareness while image recognition provides specific local verification - reducing false positives when the vehicle departs from the road for legitimate reasons like parking

Inventive Principle:
Principle #1Segmentation

2Measurement precision

If parking frame line candidates are extracted from bird's-eye view images, then parking space detection accuracy is improved, but processing complexity increases

Engineering Contradiction:
Improveparking frame detection accuracyVSAvoidimage processing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent transforms the two-dimensional image data into a bird's-eye view representation, adding a vertical dimension to the processing. This dimensional transformation simplifies the detection of linear features like parking frame lines by presenting them in a top-down perspective where their geometric properties are more apparent and easier to measure accurately

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Solution Approach 2:

The patent performs preliminary processing by converting images to bird's-eye view before extracting parking frame line candidates. This preliminary transformation prepares the data in an optimal format for subsequent line detection, making the extraction process more efficient and accurate while reducing the computational burden of the overall system

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP2927077B1Vehicular acceleration suppression device
Publication Date: 2017.09.06 NISSAN MOTOR CO LTD
  • EP2927077B1 patent drawingFigure 1
  • EP2927077B1 patent drawingFigure 2
  • EP2927077B1 patent drawingFigure 3

AI summary

To provide a vehicular acceleration suppression device capable of reducing the degradation in the drive performance due to the acceleration suppression control activated outside the parking region. A parking frame line candidate which is a candidate of a parking frame line is extracted from a bird's-eye view image acquired by performing a bird's-eye view conversion on a captured image around a vehicle (V), and a parking frame line candidate corresponding line corresponding the parking frame line candidate is extracted from the captured image of the front of the vehicle (V). Then, the parking frame line candidate corresponding to the parking frame line candidate corresponding line length of which is longer than a parking frame line length threshold (Lth1) is eliminated from the candidate to be detected as the parking frame.