Monocular Self-Position Estimation Using External Object Scale
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The accuracy of self-position estimation in vehicles using monocular cameras is compromised due to indefinite actual scale in captured images, and reliance on GNSS receivers in environments with insufficient positioning information degrades estimation accuracy, while storing landmark feature points increases memory burden.
Innovation Solution
A self-position estimation device utilizing a camera and communication device to acquire feature points, calculate actual scale, and estimate self-position based on object information from external devices, reducing memory storage needs and enabling accurate estimation even in environments where GNSS is unreliable.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a monocular camera is used for self-position estimation, then the device complexity is reduced, but the measurement precision of self-position degrades due to indefinite actual scale
Solution Approach 1:
The patent introduces an external device as an intermediary that provides object information (such as landmark positions and scales) to the self-position estimation device. This mediator resolves the scale ambiguity of monocular camera images by providing reference information from outside the system, enabling accurate self-position estimation without increasing the complexity of the camera system itself.
2Measurement precision
If landmark feature points are pre-stored in a database, then the measurement precision of self-position estimation is improved, but the quantity of substance (memory storage) increases
Solution Approach 1:
The external device performs self-service by autonomously providing object information to the self-position estimation device when needed. Instead of requiring the system to store extensive landmark databases, the external device generates and transmits relevant object information on-demand, reducing memory storage requirements while maintaining estimation accuracy.
Solution Approach 2:
The patent extracts the database storage function from the self-position estimation device and relocates it to an external device. Only essential object information is transmitted when needed, rather than storing complete landmark databases locally. This extraction reduces the memory burden on the estimation device while preserving measurement precision.
3Device complexity
If GNSS receiver is used in environments with insufficient positioning information, then the device complexity is reduced, but the reliability of self-position estimation degrades
Solution Approach 1:
The patent merges multiple information sources: monocular camera images, object information from external devices, and GNSS positioning information. By combining these diverse data sources, the system achieves reliable self-position estimation in environments where GNSS alone would be insufficient, without significantly increasing device complexity since the external device handles much of the information processing.
Data Source
AI summary
A self-position estimation device is adapted to a movable object having a camera and a communication device. The camera takes an image of surrounding of the movable object. The self-position estimation device includes a feature acquisition unit, a scale calculation unit, and a self-position estimation unit. The feature acquisition unit acquires an image feature point. The scale calculation unit calculates an actual scale of the image. The self-position estimation unit estimates a self-position of the movable object based on the image feature point and the actual scale. The scale calculation unit acquires object information and an object image feature point, and calculates the actual scale based on the acquired object information and a feature value of the acquired object image feature point.