Hospital Bed Occupancy Detection Using 3D Point Clouds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for detecting patient occupancy on hospital beds in operating rooms rely on RGB/color images, which raise privacy concerns and fail when patients are occluded by blankets or sheets, necessitating a method that uses depth images to accurately determine bed occupancy without RGB information.
Innovation Solution
A system utilizing depth cameras to generate 3D point clouds of hospital beds, extracting geometric features, and applying a trained binary classifier to determine if a bed is occupied by a patient, operating exclusively on depth sensor outputs to ensure privacy and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If RGB/color images are used for detecting patient occupancy, then detection accuracy is improved, but privacy concerns arise and system complexity increases due to PII removal requirements
Solution Approach 1:
The patent extracts only the necessary depth information from the environment while discarding all color and texture information. By using depth cameras and point cloud processing, the system extracts geometric features (distance, shape, spatial relationships) that are sufficient for occupancy detection without containing any personally identifiable information, thus eliminating privacy concerns while maintaining detection accuracy.
Solution Approach 2:
The patent replaces the optical/mechanical RGB imaging system with a depth sensing system that uses time-of-flight or structured light measurement. This substitution changes the detection mechanism from capturing reflected visible light to measuring light travel time or phase shift, providing accurate depth information without color data that could reveal personal identity.
2Measurement precision
If RGB images are used for detecting patient occupancy, then detection accuracy is improved, but privacy protection deteriorates due to PII in captured images
Solution Approach 1:
The system extracts only depth and geometric information from the scene while deliberately excluding all color, texture, and appearance information. The point cloud representation captures spatial coordinates and distances that define patient presence and bed occupancy status without containing any visual features that could identify the patient, thus achieving both accurate detection and privacy protection.
3Device complexity
If 2D image-based detection is used, then system simplicity is maintained, but detection reliability deteriorates when patients are occluded by blankets or sheets
Solution Approach 1:
The patent transitions from 2D image processing to 3D point cloud analysis by incorporating depth information as a third dimension. This dimensional enhancement allows the system to detect the spatial extent and shape of objects, enabling it to distinguish between patients covered by blankets and actual blankets or sheets based on their three-dimensional geometric characteristics, thus maintaining reliability while keeping the system relatively simple.
4Object-affected harmful factors
If depth images are used instead of RGB images, then privacy protection is improved, but detection accuracy may deteriorate due to lack of color information
Solution Approach 1:
The patent changes the detection parameter from color-based features (RGB values, textures, appearances) to depth-based features (distance, spatial coordinates, geometric shapes). This parameter transformation enables the system to detect occupancy status through the presence and shape of objects in 3D space, maintaining high accuracy while inherently protecting privacy since depth data does not reveal personal identity.
Data Source
AI summary
Embodiments described herein provide systems and techniques for detecting hospital bed occupancy based on three-dimensional (3D) point clouds of the hospital bed extracted from depth images. In one aspect, a process for determining if a bed inside an operating room (OR) is occupied by a patient is disclosed. This process begins by receiving a 3D point cloud of a bed object within a depth image captured inside the OR. The process then segments the 3D point cloud of the bed object into a plurality of segments in both a length direction and a width direction of the bed object. Next, the process extracts a set of geometric features from the plurality of segments. The process subsequently applies a binary classifier to the set of geometric features to classify the bed object as either being occupied by a patient or not being occupied by a patient.