Vehicle Object Detection with Block-Based Person Search

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing object detection devices for vehicles have a high processing load due to the large areas in which person detection processes are performed, leading to potential delays in detecting obstacles and increased computational requirements.

Innovation Solution

The object detection device partitions the XY-plane into specific blocks corresponding to the vehicle's path, focusing the person detection process on neighboring objects within these blocks, thereby reducing the processing load and improving detection efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Area of stationary object

If person detection process is performed on a large area, then detection coverage is improved, but processing load increases

Engineering Contradiction:
Improvedetection areaVSAvoidprocessing load
Core Design Contradiction:
Area of stationary objectVSProductivity

Solution Approach 1:

The detection area is divided into multiple blocks, and person detection is performed selectively in specific blocks rather than the entire area. This segmentation allows the system to maintain comprehensive detection coverage while reducing the total processing load by focusing computational resources only on relevant regions.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different blocks are assigned different detection priorities based on their relevance to the vehicle's path. Blocks containing neighboring objects that may block the vehicle are subjected to person detection, while other blocks are excluded. This local quality approach ensures detection resources are concentrated where they are most needed.

Inventive Principle:
Principle #3Local quality

2Reliability

If person detection process is performed on a large area, then detection completeness is improved, but detection time increases

Engineering Contradiction:
Improvedetection completenessVSAvoiddetection time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The detection area is segmented into multiple blocks, enabling parallel processing of different regions. By performing person detection only in blocks containing neighboring objects, the system reduces total detection time while maintaining completeness for critical areas through selective focus.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system performs preliminary identification of neighboring objects and their spatial relationships before conducting person detection. This preliminary action allows the system to pre-determine which blocks require detailed person detection, thereby avoiding unnecessary processing time in irrelevant areas while ensuring complete detection where needed.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3800576B1Object detection device, vehicle, and object detection process
Publication Date: 2025.05.14 TOYOTA INDUSTRIES CORP
  • EP3800576B1 patent drawingFigure 1~2
  • EP3800576B1 patent drawingFigure 3
  • EP3800576B1 patent drawingFigure 4

AI summary

An object detection device (41) configured to be mounted on a vehicle (10) includes an object extraction unit that is configured to extract a point group that is a set of points representing a part of an object (O) as the object (O), a neighboring object extraction unit that is configured to extract a neighboring object (O1) in an XY-plane of a world coordinate system, wherein the neighboring object (O1) is the object (O) which is the closest to the vehicle (10), a coordinate transformation unit that is configured to transform coordinates of the neighboring object (O1) in the world coordinate system into coordinates of the neighboring object (O1) in an image (11) captured by a camera (31), and a person determination unit that is configured to perform a person detection process in which it is determined whether or not the neighboring object (O1) is a person on the coordinates of the neighboring object (O1).