Face Detection Coordinate Correction for Rotating Camera Displays

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Digital cameras with rotatable display units face challenges in face detection reliability due to inconsistencies between the image coordinates displayed on the screen and those detected by the image sensor, especially when the camera is rotated, leading to inaccurate or failed face detection.

Innovation Solution

A subject detecting method and apparatus that corrects face detection coordinates by rotating and inverting the input image based on the rotation state of the camera, ensuring the detected face coordinates align with the displayed image, using a combination of acceleration and Hall sensors to sense the rotation and display states, and a digital signal processor to perform image transformations and display adjustments.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If the display unit is made rotatable to improve photographing convenience, then the ease of operation is improved, but the reliability of face detection deteriorates due to coordinate inconsistency between the sensor and display

Engineering Contradiction:
Improvephotographing convenienceVSAvoidface detection reliability
Core Design Contradiction:
Ease of operationVSReliability

Solution Approach 1:

The system performs preliminary coordinate transformation based on detected rotation angles before face detection. The processor calculates transformed coordinates using rotation matrices and applies these transformations to align the display coordinate system with the sensor coordinate system, ensuring accurate face detection regardless of camera orientation

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system uses acceleration sensors and Hall sensors to continuously detect the camera's rotation state and provides feedback to the processor. The processor dynamically adjusts coordinate transformations based on this feedback, maintaining alignment between display and sensor coordinates even during rotation

Inventive Principle:
Principle #23Feedback

2Reliability

If coordinate transformation is applied to align display and sensor coordinates, then the reliability of face detection is improved, but the device complexity increases due to additional processing steps

Engineering Contradiction:
Improveface detection reliabilityVSAvoidprocessing complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The system changes coordinate parameters dynamically based on rotation angles. The processor stores pre-calculated rotation matrices for different angles (0°, 90°, 180°, 270°) and selects the appropriate transformation, avoiding complex real-time calculations while maintaining accuracy

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If multiple sensors are added to detect rotation state, then the measurement precision of rotation angle is improved, but the device complexity increases

Engineering Contradiction:
Improverotation angle detection precisionVSAvoidsensor system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system combines data from acceleration sensors and Hall sensors to detect rotation state. The processor integrates information from both sensor types, using their complementary strengths to achieve accurate orientation detection while maintaining system compactness

Inventive Principle:
Principle #5Merging (Combining)

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

Enhances the reliability of face detection in digital cameras by aligning detected face coordinates with the displayed image, even when the camera is rotated, allowing for efficient and accurate subject detection without requiring additional learning methods.

Implementation Method 1

a rotation state of the input image and a display state of the input image may be ascertained

Methodology Applied
Scientific EffectGravitation: Gravitation

Implementation Method 2

using a combination of acceleration and Hall sensors to sense the rotation and display states

Methodology Applied
Scientific EffectHall Effect: Hall Effect

Data Source

PatentUS9111129B2Subject detecting method and apparatus, and digital photographing apparatus
Publication Date: 2015.08.18 SAMSUNG ELECTRONICS CO LTD
  • US9111129B2 patent drawing
  • US9111129B2 patent drawing
  • US9111129B2 patent drawing

AI summary

A subject detecting method and a subject detecting apparatus, by which face detection may be efficiently performed in a digital photographing apparatus having a flippable display unit, and when an image input via an image sensor of the digital photographing apparatus is different from an image displayed on the display unit due to rotation of the digital photographing apparatus, a face detection coordinate may be corrected to increase the reliability of face detection.