Region-Focused Image Recognition for High-Precision Road Sign Reading

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image capturing systems in autonomous vehicles face challenges in efficiently recognizing road signs and other objects with high precision due to high data amounts and difficulties in smooth data transmission when increasing image resolution.

Innovation Solution

A recognition apparatus that includes a detection unit to identify regions of interest, a control unit to generate focused images of those regions, and an image processing unit to enhance resolution, allowing for precise content recognition while reducing data volume.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the resolution of captured images is increased to recognize road signs with high precision, then recognition precision is improved, but data amount becomes high making smooth transmission difficult

Engineering Contradiction:
Improverecognition precisionVSAvoiddata amount
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent divides the image processing into two stages: first capturing a low-resolution wide-area image to detect the subject's location, then capturing a high-resolution image only of the determined region containing the subject. This segmentation of processing scope reduces overall data amount while maintaining recognition precision for the target object.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies different quality levels to different regions of the captured scene. The determined region containing the subject is captured at high resolution for precise recognition, while the surrounding areas are captured at low resolution or not captured at all, optimizing the balance between recognition precision and data transmission efficiency.

Inventive Principle:
Principle #3Local quality

2Measurement precision

If high-resolution images are captured for all regions, then recognition precision is improved, but transmission efficiency deteriorates

Engineering Contradiction:
Improverecognition precisionVSAvoidtransmission efficiency
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent extracts only the necessary region containing the subject from the full scene for high-resolution processing. By using detection results to determine the subject's location and extracting only this region for enhanced resolution, the system maintains recognition precision while significantly reducing the volume of high-resolution data that needs to be transmitted.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent performs preliminary detection and region determination on low-resolution images before capturing high-resolution images. This preliminary action identifies the exact area requiring high precision, allowing the system to prepare the appropriate capture parameters and minimize unnecessary high-resolution data generation, thereby improving transmission efficiency.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS12633134B2Recognition apparatus, recognition method, and storage medium
Publication Date: 2026.05.19 CANON KK
  • US12633134B2 patent drawing
  • US12633134B2 patent drawing
  • US12633134B2 patent drawing

AI summary

A recognition apparatus including: a detection unit configured to detect a subject to be recognized from a first image that has been captured by an image capturing apparatus; a determining unit configured to, in a case in which the subject has been detected from the first image, determine a region including at least a portion of the subject in the first image; a control unit configured to control the image capturing apparatus so as to generate a second image that corresponds to the region; an image processing unit configured to generate a third image by processing to increase a resolution of the second images; and a recognition unit configured to recognize contents that are indicated by the subject from the third image.