Road Mirror Object Detection via Dual-Resolution Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing object detection systems for self-driving cars fail to efficiently detect moving objects reflected in road mirrors, leading to increased processing demands and potential real-time detection failures, especially when resources are limited.

Innovation Solution

An object detection device comprising a first detecting part that uses low-resolution image data to identify road mirrors and potential reflections, and a second detecting part that utilizes high-resolution partial image data to precisely locate moving objects within the road mirror, reducing overall processing load while maintaining accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If high resolution image data is used to detect moving objects in road mirrors, then detection precision is improved, but processing load increases

Engineering Contradiction:
Improvedetection precisionVSAvoidprocessing load
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent divides the detection process into two segments: first detecting road mirrors and their reflected objects using low-resolution images, then selectively extracting and processing only the relevant road mirror regions at high resolution. This segmentation allows precise detection of moving objects in road mirrors while avoiding the need to process entire high-resolution images, thus reducing overall processing load.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies different image resolutions to different detection tasks: low-resolution images are used for initial scene understanding and road mirror detection, while high-resolution images are used only for the specific local region of road mirrors where precise moving object detection is needed. This local quality approach optimizes the balance between detection precision and processing efficiency.

Inventive Principle:
Principle #3Local quality

2Productivity

If processing resources are limited or real-time detection is required, then processing speed is improved, but detection capability for reflected objects deteriorates

Engineering Contradiction:
Improvedetection speedVSAvoiddetection capability
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent performs preliminary detection of road mirrors and their reflected objects using low-resolution images before conducting detailed moving object detection. This preliminary action identifies which regions require high-resolution processing, enabling real-time detection by avoiding unnecessary high-resolution processing of entire scenes while maintaining reliable detection capability for reflected objects.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS11423663B2Object detection device, object detection method, and non-transitory computer readable medium comprising computer program for object detection-use
Publication Date: 2022.08.23 TOYOTA JIDOSHA KK
  • US11423663B2 patent drawing
  • US11423663B2 patent drawing
  • US11423663B2 patent drawing

AI summary

An object detection device comprising a first detecting part detecting types and positions of objects including a road mirror in an image and a presence of a movable object reflected in the road mirror based on image data of a low resolution image and a second detecting part detecting a type of the movable object in the road mirror and a position of the movable object in the road mirror based on high resolution partial image data obtained by extracting the part of the road mirror from the image if there is a movable object reflected in the road mirror.