Mobile Camera Frame Selection with Optical Flow Motion Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Images captured by mobile devices with motion are often of poor quality, which existing technologies fail to effectively identify and improve.

Innovation Solution

An image processing method that extracts feature points using algorithms like SIFT or SURF, tracks optical flow, calculates displacement distances, and determines valid frame images based on feature point quantity, definition, and motion state to output high-quality images.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If mobile device captures images during motion, then photography convenience is improved, but image quality deteriorates

Engineering Contradiction:
Improvephotography convenienceVSAvoidimage quality
Core Design Contradiction:
Ease of operationVSManufacturing precision

Solution Approach 1:

The system performs preliminary actions by continuously capturing preview frames before the user takes the final photo, and pre-processes these frames to identify motion blur. This allows the system to have ready-made candidate frames that are less likely to be blurred, resolving the contradiction by preparing quality alternatives in advance while maintaining the convenience of motion photography

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system implements feedback by analyzing motion blur in captured frames and using this information to select or reject frames. The motion blur detection algorithm provides feedback about image quality, enabling the system to automatically choose frames with better quality even when the device is moving, thus maintaining both convenience and quality

Inventive Principle:
Principle #23Feedback

2Reliability

If all captured images are processed and output, then image selection completeness is improved, but processing efficiency deteriorates

Engineering Contradiction:
Improveimage selection completenessVSAvoidprocessing efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The system extracts only the necessary information from captured frames - specifically motion blur characteristics and quality metrics - rather than processing entire images. By extracting key features and using them to filter frames, the system maintains selection completeness while dramatically reducing processing requirements and improving efficiency

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The system performs partial processing on multiple frames by analyzing only critical quality indicators rather than fully processing each image. This partial action approach ensures that the best frames are selected (completeness) while avoiding the computational overhead of processing every frame in full detail (efficiency)

Inventive Principle:
Principle #16Partial or excessive action

3Measurement precision

If motion blur detection is implemented, then image quality identification is improved, but computational complexity increases

Engineering Contradiction:
Improveimage quality identificationVSAvoidcomputational complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system replaces complex mechanical or manual quality assessment with automated computational algorithms that detect motion blur through image processing. These algorithms analyze pixel patterns and motion characteristics to identify blur, providing precise quality identification while keeping computational requirements manageable through efficient algorithm design

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentEP3657436B1Image processing method, terminal, and storage medium
Publication Date: 2025.08.20 TENCENT TECHNOLOGY (SHENZHEN) CO LTD
  • EP3657436B1 patent drawingFigure 1A
  • EP3657436B1 patent drawingFigure 1B~2A
  • EP3657436B1 patent drawingFigure 2B

AI summary

An image processing method is provided, including: extracting at least one image feature point from a target frame image captured by a camera; obtaining feature information of at least one optical flow image feature point in a previous frame image captured by the camera in a case that the target frame image is a valid frame image, and determining, among the at least one image feature point, an image feature point corresponding to the optical flow image feature point according to feature information of the at least one image feature point; calculating a displacement distance, in a frame image, between the optical flow image feature point in the previous frame image and the corresponding image feature point in the target frame image; and outputting the target frame image in response to a determination that the target frame image is captured in a non-motion state according to the displacement distance.