Image Orientation Detection Using Shape Pattern Matching

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image processing techniques are limited in determining the correct orientation of images with various subjects and compositions, as they rely on color features and face alignment, which are not robust to light source environments and are applicable only to images with multiple people captured side by side.

Innovation Solution

An image processing device and method that detects objects by pattern matching between an image and a reference image while changing the relative angle, identifying the top and bottom of the image to determine its correct orientation, and updates the image data for proper display.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If skin area detection and alignment-based orientation determination is used, then the vertical direction can be determined for images with multiple people, but the method is weak in robustness to light source environment, color of clothes, and positions of arms and legs

Engineering Contradiction:
Improverobustness to light source environment and color variationsVSAvoidapplicability to various subjects and compositions
Core Design Contradiction:
ReliabilityVSAdaptability or versatility

Solution Approach 1:

The patent changes the detection parameter from color-based skin area detection to shape-based pattern matching. By using template matching with predefined object shapes (faces, bodies, objects) and comparing geometric patterns rather than color features, the method achieves robustness against variations in lighting conditions, clothing colors, and body positions while maintaining applicability to diverse image content.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If face alignment-based orientation determination is used, then the vertical direction can be determined, but it is assumed that a plurality of people are being captured side by side, thus limiting applicability

Engineering Contradiction:
Improveaccuracy of orientation determinationVSAvoidapplicability to images with various subjects
Core Design Contradiction:
Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent creates a universal orientation determination system that can handle multiple types of subjects (people, objects, landscapes) through a single pattern matching framework. By defining multiple template patterns for different object types and using shape-based rather than face-specific detection, the system achieves multi-functionality across diverse image content while maintaining precise orientation determination through the identified object's alignment.

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

3Adaptability or versatility

If pattern matching with angle changing is performed, then objects can be detected in various orientations, but the processing time increases

Engineering Contradiction:
Improvedetection capability for various orientationsVSAvoidprocessing time for orientation detection
Core Design Contradiction:
Adaptability or versatilityVSLoss of time

Solution Approach 1:

The patent applies partial action by performing pattern matching at selectively chosen angles rather than exhaustively testing all possible orientations. The system rotates the image or template at discrete angle intervals (e.g., 0°, 45°, 90°, 135°) to detect object orientation, achieving sufficient adaptability for various orientations while significantly reducing processing time compared to continuous angle scanning.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentUS9049397B2Image processing device and image processing method
Publication Date: 2015.06.02 SONY INTERACTIVE ENTERTAINMENT LLC
  • US9049397B2 patent drawing
  • US9049397B2 patent drawing
  • US9049397B2 patent drawing

AI summary

A face detection processing unit performs a face detection process by rotating an image in increments of a predetermined angle and acquires a rotation angle at which a face is detected. An angle correction unit acquires an angle between the face and a shoulder by pattern matching and corrects the rotation angle of the image. A human-image orientation identification unit identifies a correct orientation of a human image based on the rotation angle. An image-characteristic analysis unit analyzes a frequency distribution and a brightness distribution of a non-human image. A non-human image orientation identification unit identifies the correct orientation of a non-human image based on distribution characteristics of a frequency of brightness with respect to an axis in a predetermined direction. An image-data updating unit incorporates information regarding the correct orientation in the image data.