Wearable Obstacle Detection Using Orientation-Corrected Normal Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing visual aid devices for low vision individuals are not sufficiently accurate and versatile in detecting obstacles, leading to mobility issues and limited autonomy due to false detections.
Innovation Solution
A method using a distance sensor mounted on a wearable device to obtain a normal map, apply orientation correction, and segment the map into zones like floor, obstacle, or lateral wall, employing techniques such as ToF, LIDAR, or RGB imaging, with orientation correction using IMU or visual odometry to account for head movements, and threshold comparisons to classify surfaces accurately.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional distance sensors (ultrasound, simple ToF) are used for obstacle detection, then the device is cost-effective and simple, but the detection accuracy is insufficient and generates too many false detections
Solution Approach 1:
The patent combines multiple sensor types (ToF sensor, LIDAR, or RGB-D camera) with IMU sensors and integrates them into a unified processing system. The ToF/LIDAR/RGB-D sensor provides depth information while the IMU provides orientation data, and their combined processing through coordinate system transformation and normal map analysis resolves the contradiction by achieving high accuracy without requiring overly complex individual sensor components
Solution Approach 2:
The system uses a multi-functional approach where the distance sensor can operate in multiple modes (ToF, LIDAR, or RGB-D imaging) and the processing system performs multiple functions including coordinate transformation, normal map generation, orientation correction, and zone segmentation. This universality allows the system to maintain high detection accuracy while adapting to different sensor configurations without proportionally increasing complexity
2Ease of operation
If the distance sensor is mounted on a wearable device to enable mobility, then the system becomes portable and useful for low vision individuals, but the sensor orientation changes with head movements causing detection errors
Solution Approach 1:
The system continuously monitors the sensor orientation through IMU data and dynamically adjusts the coordinate system transformation based on the detected head position and orientation. This feedback mechanism ensures that even as the wearable device moves with the user's head, the system compensates for orientation changes in real-time, maintaining detection accuracy while preserving portability
Solution Approach 2:
The patent implements a dynamic coordinate system transformation that adapts to changing sensor orientations. Instead of using a fixed reference frame, the system continuously transforms sensor data into the world coordinate system based on real-time IMU measurements of head position and orientation. This dynamic adaptation allows the wearable system to maintain measurement precision despite constant movement and orientation changes
3Measurement precision
If normal map processing with orientation correction is applied, then the system achieves accurate obstacle segmentation independent of head movements, but the processing complexity and computational load increase
Solution Approach 1:
The system performs preliminary coordinate system transformation and orientation correction on the normal map data before conducting obstacle segmentation. By pre-aligning the sensor data to the world coordinate system using IMU-based orientation information, the system simplifies subsequent processing steps and reduces the computational complexity of obstacle detection, achieving high accuracy without proportionally increasing overall processing complexity
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The method provides a robust and reliable detection of obstacles, reducing false classifications and enhancing mobility by accurately segmenting the environment into relevant zones, independent of user head movements.
Implementation Method 1
ToF (Time of Flight) based sensors: a laser beam is sent to the target and the time elapse in between the sending and the receiving of the bounced beam is measured. This way, by having the speed of light and the time of arrival of the beam, distance can be measured.
Implementation Method 2
LIDAR sensors (which are precise and fast and have a long range)
Implementation Method 3
Modulated light scanner based sensors: They are similar to ToF sensors. In this case a modulated light signal is sent to the target and the difference in between the sent signal and the received signal (which has bounced on the target) is calculated, in order to obtain the distance to the target.
Data Source
AI summary
Detecting obstacle elements, the detection performed with a visual aid device, the device including a distance sensor, the sensor being mounted in a wearable device to be worn by a user. Several embodiments allow the further detection of different types of obstacles, in order to map and signal them to a user with low-vision.