Infant Monitoring ROI Detection for Lower Processing Demand

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image-based infant monitoring systems require expensive technology and high computational power to accurately track infant behavior due to the need for processing large amounts of image data, leading to inefficiencies and potential false positives.

Innovation Solution

A method is proposed to determine a region of interest within an image using feature detection models to identify the infant's body, allowing for focused processing and analysis on this region, reducing computational requirements and improving accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the entire image data is processed to monitor the infant, then the monitoring accuracy is improved, but the computational power requirement increases

Engineering Contradiction:
Improvemonitoring accuracyVSAvoidcomputational power
Core Design Contradiction:
Measurement precisionVSPower

Solution Approach 1:

The patent divides the entire image into multiple regions of interest (ROIs) based on detecting infant body parts such as head, torso, and limbs. Instead of processing the complete image, the system segments and processes only the relevant regions where the infant is present, thereby reducing computational power requirements while maintaining monitoring accuracy.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent extracts and processes only the essential portions of the image data corresponding to the infant's body regions. By identifying and isolating these specific regions through feature detection, the system removes unnecessary background data and focuses computational resources only on the extracted infant-related regions, reducing overall computational power consumption.

Inventive Principle:
Principle #2Taking out (Extraction)

2Reliability

If the entire image data is processed to monitor the infant, then the monitoring coverage is improved, but the processing time increases

Engineering Contradiction:
Improvemonitoring coverageVSAvoidprocessing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent segments the image into multiple regions of interest corresponding to different body parts of the infant. This segmentation allows the system to process only the relevant regions simultaneously, maintaining comprehensive monitoring coverage while significantly reducing the total processing time compared to analyzing the entire image.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies partial action by processing only the necessary portions of the image data (the regions containing the infant) rather than the entire image. This selective processing approach ensures adequate monitoring coverage of the infant while reducing processing time by excluding irrelevant background areas.

Inventive Principle:
Principle #16Partial or excessive action

3Measurement precision

If feature detection models are used to identify the infant, then the monitoring accuracy is improved, but the device complexity increases

Engineering Contradiction:
Improvemonitoring accuracyVSAvoiddevice complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent uses pre-trained feature detection models that have been copied and adapted for infant monitoring. Instead of developing new complex detection algorithms from scratch, the system leverages existing models (such as those trained on general object detection) and fine-tunes them for infant-specific features, thereby improving accuracy while avoiding the complexity of creating new detection systems.

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The patent employs universal feature detection models that can identify various infant body parts (head, torso, limbs) using a single integrated system. This multi-functional approach allows one model to handle multiple detection tasks, improving monitoring accuracy without proportionally increasing device complexity.

Inventive Principle:
Principle #6Universality (Multi-functionality)

4Power

If the region of interest is reduced, then the computational requirements are reduced, but the monitoring accuracy may deteriorate

Engineering Contradiction:
Improvecomputational requirementsVSAvoidmonitoring accuracy
Core Design Contradiction:
PowerVSMeasurement precision

Solution Approach 1:

The patent applies local quality by focusing computational resources on specific regions of interest where the infant is detected. Each region is processed with appropriate detail and quality tailored to its importance, ensuring that monitoring accuracy is maintained in the critical infant areas while reducing computational requirements in the broader context of the entire image.

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP4625361A1Infant monitoring
Publication Date: 2025.10.01 KONINKLIJKE PHILIPS NV
  • EP4625361A1 patent drawingFigure 1
  • EP4625361A1 patent drawingFigure 2
  • EP4625361A1 patent drawingFigure 3

AI summary

Proposed concepts aim to provide schemes, solutions, concepts, designs, methods, and systems pertaining to determining a region of interest for use in monitoring an infant. In particular, embodiments aim to provide a method for determining a region of interest for use in monitoring an infant by processing an image of an area using a feature detection model to detect a feature of the infant's body. Based on the position of the detected feature within the area, a region of interest within the area is then determined. Such a region of interest may help to identify image data than can be ignored or deprioritized by infant one or more monitoring processes, thus reducing an amount of processing/analysis required to be undertaken by the process(es).