Image Frame Integration with Motion-Aware Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image frame integration methods are often degraded by motion blur and noise, failing to effectively distinguish between stationary and moving objects, which leads to erroneous motion detection and blurring of moving objects.

Innovation Solution

The system integrates image frames by calculating an average difference in sensor measurements across multiple frames, using a persistent count to identify motion, and applying a noise filter to reduce false detections, thereby reducing motion blur and enhancing image clarity.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If frame integration is applied to reduce noise, then image quality improves, but motion blur increases

Engineering Contradiction:
Improveimage qualityVSAvoidmotion blur
Core Design Contradiction:
ReliabilityVSObject-affected harmful factors

Solution Approach 1:

The image frame is segmented into stationary and moving regions based on motion detection. Frame integration is applied selectively: stationary regions undergo full integration to reduce noise, while moving regions are excluded from integration to prevent motion blur. This segmentation approach allows simultaneous noise reduction and motion preservation.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different processing quality is applied to different regions of the image. Stationary regions receive high-quality integrated processing with full noise reduction, while moving regions receive lower-quality processing that preserves temporal sharpness. This local quality differentiation resolves the contradiction by optimizing each region for its specific characteristics.

Inventive Principle:
Principle #3Local quality

2Measurement precision

If motion detection threshold is lowered to detect more motion, then moving objects are better identified, but false detections increase

Engineering Contradiction:
Improvemotion detection accuracyVSAvoidfalse detection rate
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

A persistent count mechanism is used to preliminarily evaluate motion persistence before final detection. When pixel differences exceed the threshold, the system increments a counter that must reach a predetermined value before confirming motion. This preliminary counting action filters out transient noise while preserving genuine motion detection.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system uses feedback through the persistent count mechanism where past motion detection results influence current detection decisions. The counter accumulates evidence of motion over multiple frames, and this accumulated feedback determines whether motion is confirmed. This feedback loop reduces false detections while maintaining sensitivity to genuine motion.

Inventive Principle:
Principle #23Feedback

3Object-affected harmful factors

If frame integration is disabled for moving pixels, then motion blur is reduced, but noise reduction effectiveness decreases

Engineering Contradiction:
Improvemotion blurVSAvoidnoise reduction
Core Design Contradiction:
Object-affected harmful factorsVSReliability

Solution Approach 1:

The image is segmented into stationary and moving regions, allowing differential processing. Stationary regions undergo full frame integration for maximum noise reduction, while moving regions are excluded from integration to prevent blur. This segmentation ensures optimal noise reduction where applicable without compromising motion regions.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different integration quality is applied locally: high-quality full integration for stationary regions and low-quality or no integration for moving regions. This local differentiation maximizes noise reduction in appropriate areas while preserving motion sharpness where needed.

Inventive Principle:
Principle #3Local quality

4Measurement precision

If pixel-by-pixel comparison is used for motion detection, then detection precision improves, but computational complexity increases

Engineering Contradiction:
Improvemotion detection precisionVSAvoidcomputational complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system performs pixel-by-pixel comparison only where necessary - specifically for pixels that may be part of moving objects or at region boundaries. Once motion is detected in a region, surrounding pixels are assumed to share the same motion status, reducing the need for exhaustive pixel-by-pixel comparison across the entire image. This partial action maintains precision while reducing complexity.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP3055988B1Integrating image frames
Publication Date: 2021.05.19 RAYTHEON CO
  • EP3055988B1 patent drawingFigure 1
  • EP3055988B1 patent drawingFigure 2
  • EP3055988B1 patent drawingFigure 3

AI summary

A method for integrating image frames includes receiving for each pixel in at least one image frame a value representative of a sensor measurement. The method includes calculating an average difference of the value representative of the sensor measurement over a subset of the plurality of image frames. The method includes detecting motion in at least one pixel of the image frame in the plurality of image frames based on the calculated average difference of the value representative of the sensor measurement over the subset of the plurality of image frames. The method also includes generating an integrated image frame wherein each pixel having detected motion is integrated by an amount less than that of those pixels for which motion is not detected. The amount of frame integration is based on contrast levels, expected rates of motion, and noise characteristics of sensor input image data.