Image Processing Device Two-Stage Part Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image processing technologies often fail to detect specific parts of objects, such as license plates or faces, even when the object itself is detected, particularly when the object's parts are small in relation to the entire object or when image resolution is reduced to decrease processing load.

Innovation Solution

An image processing device comprising a first detection unit to detect objects and their parts, a setting unit to adjust the detection range based on the first detection unit's results, and a second detection unit to specifically detect the parts within the set range, using machine-learned models to enhance detection accuracy and efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If the resolution of the captured image is reduced to decrease processing load, then processing load is reduced, but detection accuracy of object parts deteriorates

Engineering Contradiction:
Improveprocessing loadVSAvoiddetection accuracy of object parts
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The detection process is divided into two stages: first detecting the entire object, then separately detecting the object part within a focused range. This segmentation allows the system to maintain high detection accuracy for parts while processing the entire image at a lower effective resolution, thus reducing overall processing load without sacrificing part detection capability.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system applies different detection strategies to different regions of the image. The object detection is performed across the entire image, while the object part detection is concentrated in a localized range around the detected object. This local quality approach ensures high detection accuracy for parts where it matters most while reducing processing requirements for the overall system.

Inventive Principle:
Principle #3Local quality

2Reliability

If the detection range is expanded to detect small object parts, then detection coverage is improved, but processing complexity increases

Engineering Contradiction:
Improvedetection coverageVSAvoidprocessing complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The system performs preliminary detection of the entire object before attempting to detect the object part. By first identifying the object's location and boundaries, the system establishes a predetermined range that guides the subsequent part detection process. This preliminary action reduces processing complexity by avoiding the need to search the entire image for small parts, thereby maintaining high detection coverage without excessive complexity.

Inventive Principle:
Principle #10Preliminary action

3Reliability

If multiple detection processes are implemented to improve part detection, then detection reliability is improved, but processing time increases

Engineering Contradiction:
Improvepart detection reliabilityVSAvoidprocessing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The detection system dynamically adjusts the detection range based on the detected object's characteristics. After detecting the object, the system determines a predetermined range around the object and focuses the part detection process within this dynamic range rather than searching the entire image. This dynamic approach maintains high part detection reliability while significantly reducing processing time compared to exhaustive full-image searches.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentUS12299987B2Image processing device and image processing method
Publication Date: 2025.05.13 DENSO TEN LTD
  • US12299987B2 patent drawing
  • US12299987B2 patent drawing
  • US12299987B2 patent drawing

AI summary

An object to reduce a possibility that a detection target appearing in a captured image is not detected. In order to achieve such an object, an image processing device includes: a first detection unit that is configured to detect a predetermined object and a predetermined part of the predetermined object from an acquired captured image; a setting unit that sets a detection range in the captured image based on a detection result of the first detection unit in a case where only the predetermined object out of the predetermined object and the predetermined part has been detected by the first detection unit; and a second detection unit that attempts to detect the predetermined part in the detection range.