Multi-Angle Surface Identification for Similar-Looking Objects

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing technologies struggle to accurately identify the surface of an object facing an imaging device when the object has multiple surfaces with similar appearances, leading to ambiguity in orientation determination.

Innovation Solution

A processing device and program that utilize a controller to obtain multiple images of the object in different orientations, perform orientation estimation processes using template matching, and rotate the object to capture images from different angles to distinguish between similar surfaces, thereby identifying the target surface facing the imaging device.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If template matching is used to identify surface orientation, then processing speed is improved, but measurement precision deteriorates when multiple surfaces have similar appearances

Engineering Contradiction:
Improveprocessing speedVSAvoidsurface identification accuracy
Core Design Contradiction:
SpeedVSMeasurement precision

Solution Approach 1:

The system performs preliminary rotation of the object to specific orientations before imaging. By pre-positioning the object in known orientations (e.g., rotating by predetermined angles), the system creates distinguishable visual patterns that enable accurate surface identification through template matching, resolving the ambiguity of similar appearances.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system transitions from a single 2D image to a multi-angle view by rotating the object in 3D space. This dimensional change provides additional visual information that distinguishes between surfaces with similar appearances, allowing the template matching algorithm to accurately identify the target surface orientation.

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

2Measurement precision

If the object is rotated to capture images from different angles, then surface identification accuracy is improved, but device complexity increases

Engineering Contradiction:
Improvesurface identification accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The object itself is rotated to provide the necessary view changes, rather than moving the imaging device or using multiple fixed cameras. This self-rotation approach simplifies the overall system architecture while achieving multi-angle imaging, as only the object positioning mechanism needs to be controlled.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

A single imaging device captures all necessary images by combining multiple object orientations. The same camera and processing system handle both the rotation control and image analysis, making the system multi-functional and reducing the need for separate specialized equipment for each function.

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Measurement precision

If multiple images in different orientations are obtained, then surface identification accuracy is improved, but loss of time increases

Engineering Contradiction:
Improvesurface identification accuracyVSAvoidtime for multiple imaging operations
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The object is rotated to specific predetermined orientations and imaged in a systematic sequence. By using periodic rotation at fixed angles (e.g., 0°, 45°, 90°), the system efficiently captures all necessary views in a regular pattern, minimizing the total time required compared to random or continuous rotation approaches.

Inventive Principle:
Principle #19Periodic action

Data Source

PatentEP4660946A1Processing device and program
Publication Date: 2025.12.10 KYOCERA CORP
  • EP4660946A1 patent drawingFigure 1~2
  • EP4660946A1 patent drawingFigure 3
  • EP4660946A1 patent drawingFigure 4

AI summary

A processing device includes a controller that (s2) obtains, from an imaging device, a first image of an object in a first orientation. The object includes a plurality of surfaces having appearances similar to each other. The controller (s3) identifies, from the plurality of surfaces, a target surface facing the imaging device. The controller further (s5) obtains a second image of the object in a second orientation different from the first orientation. The controller (s9) identifies the target surface based on the first image and the second image.