Sensor-Aided Spatio-Temporal Feature Extraction for Video Stabilization

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for extracting spatio-temporal features from user-generated videos are ineffective due to camera motion, which corrupts the motion information of recorded objects or scenes, making it difficult to analyze videos captured by uncontrolled devices like smartphones.

Innovation Solution

A method that utilizes auxiliary sensor data from devices like accelerometers and gyroscopes to segment video frames based on camera view motion, isolating 'stable segments' with minimal camera motion and modifying feature descriptors to account for camera motion, ensuring that extracted features represent real object motion independently of camera movement.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If temporal features are extracted from user-generated videos, then motion information of objects can be captured, but camera motion corrupts the motion information making extraction ineffective

Engineering Contradiction:
Improvemotion information accuracyVSAvoidfeature extraction effectiveness
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The video is segmented into stable segments and unstable segments based on camera motion detection. Temporal features are extracted only from stable segments where camera motion is minimal, ensuring accurate motion information capture while avoiding corruption from camera movement.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Sensor data from accelerometers and gyroscopes are used as intermediary information to detect and characterize camera motion. This intermediary data allows the system to identify and compensate for camera movement effects, separating true object motion from camera-induced motion artifacts.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If video stabilization using gyroscopes is applied, then camera motion can be corrected, but device complexity increases

Engineering Contradiction:
Improvecamera motion correction accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system uses the device's own built-in sensors (accelerometer and gyroscope) that are already present in smartphones to detect and correct camera motion. This self-service approach avoids adding external stabilization equipment, maintaining device simplicity while achieving effective motion correction.

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP2763077B1Method and apparatus for sensor aided extraction of spatio-temporal features
Publication Date: 2023.11.15 NOKIA TECHNOLOGIES OY
  • EP2763077B1 patent drawingFigure 1
  • EP2763077B1 patent drawingFigure 2
  • EP2763077B1 patent drawingFigure 3

AI summary

A method, apparatus and computer program product are provided for extracting spatio-temporal features with the aid of sensor information. An exemplary method comprises receiving video data and auxiliary sensor data and associating the two with timestamp information. The method may also include segmenting an input data stream into stable segments and extracting temporal features from the associated video data. The method may further include extracting temporal features either form the whole video or only from the video data where little or no stable segments are detected and performing camera view motion compensation by using information provided by the auxiliary sensors for modifying the feature-descriptors.