Stereo Self-Position Estimation With Staggered Camera Frame Rates

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Stereoscopic SLAM systems face increased power consumption and data processing demands due to the use of two cameras, which complicates autonomous spatial movement and AR/VR applications.

Innovation Solution

A self-position estimation device that controls the imaging timing of two cameras to operate at different frame rates, allowing for reduced power consumption and data processing by estimating position using image frames captured at the same time and at different times, with an imaging control unit managing the frame rates to achieve a higher estimation rate while lowering individual camera frame rates.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If two cameras are used for stereoscopic SLAM, then self-position estimation accuracy is improved, but power consumption and data processing amount increase

Engineering Contradiction:
Improveself-position estimation accuracyVSAvoidpower consumption
Core Design Contradiction:
Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The patent implements periodic imaging with two cameras at different frame rates, where imaging is performed alternately at different time intervals. This periodic action allows the system to maintain stereoscopic SLAM functionality while reducing the overall imaging frequency and power consumption of individual cameras.

Inventive Principle:
Principle #19Periodic action

Solution Approach 2:

The patent dynamically adjusts the frame rates of the two cameras differently, allowing flexible control of imaging timing. This dynamic adjustment enables the system to optimize power consumption while maintaining accurate self-position estimation through coordinated imaging at varying rates.

Inventive Principle:
Principle #15Dynamics

2Productivity

If two cameras operate at high frame rates for stereoscopic SLAM, then estimation rate is improved, but data processing amount increases

Engineering Contradiction:
Improveestimation rateVSAvoiddata processing amount
Core Design Contradiction:
ProductivityVSQuantity of substance

Solution Approach 1:

By implementing periodic imaging at different frame rates for the two cameras, the system reduces the total number of image frames that need to be processed simultaneously, thereby reducing data processing amount while maintaining the estimation rate through coordinated timing.

Inventive Principle:
Principle #19Periodic action

Solution Approach 2:

The patent dynamically controls the frame rates of individual cameras based on system requirements, allowing the estimation rate to be maintained while reducing the overall data processing burden by adjusting imaging frequencies adaptively.

Inventive Principle:
Principle #15Dynamics

3Measurement precision

If two cameras image at the same time, then stereoscopic depth information is improved, but power consumption increases

Engineering Contradiction:
Improvestereoscopic depth informationVSAvoidpower consumption
Core Design Contradiction:
Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The patent implements periodic synchronized imaging where both cameras capture images at the same time intervals, providing stereoscopic depth information. Between these synchronized moments, cameras operate alternately at different frame rates, reducing overall power consumption while maintaining depth estimation capability.

Inventive Principle:
Principle #19Periodic action

Data Source

PatentUS11386574B2Self-position estimation device, self-position estimation method, and program thereof
Publication Date: 2022.07.12 SONY GROUP CORP
  • US11386574B2 patent drawing
  • US11386574B2 patent drawing
  • US11386574B2 patent drawing

AI summary

A self-position estimation device includes: a position estimation unit. The position estimation unit is configured to estimate a self-position on the basis of image frames that have been captured at the same time in a constant period by imaging units, and estimate a self-position on the basis of image frames that have been captured at different times in the constant period by at least one of the imaging units.