Autonomous Livestock Identification Through Visual–RFID Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing animal monitoring systems face challenges such as non-scalability, inaccuracy, timing issues, and difficulty in operating outdoors or in adverse conditions, making them inefficient and unreliable for identifying individual animals.
Innovation Solution
A method and system that uses multiple measurement modalities, including two-dimensional imaging sensors and RFID readers, to dynamically identify animals by correlating visual and RFID data, adapting to environmental changes, and learning from operating data to improve identification accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If RFID tags are used for animal identification, then identification can be obtained, but the system becomes non-scalable and prone to inconsistency and inaccuracy
Solution Approach 1:
The system segments the identification task into multiple independent components: visual identification using imaging sensors, RFID identification using readers, and sensor-based identification. Each component operates independently and can be scaled separately, allowing the overall system to be more scalable while maintaining accuracy through multiple pathways.
Solution Approach 2:
The system changes the identification parameters by using multiple modalities (visual features, RFID signals, sensor data) instead of relying solely on RFID tags. This allows the system to adapt to different situations and scale independently of any single identification method.
2Measurement precision
If specialised 3D cameras are used for animal identification, then identification accuracy can be improved, but the system becomes difficult to configure, maintain, and update
Solution Approach 1:
The system uses standard two-dimensional imaging sensors that can serve multiple functions: animal identification, monitoring, and documentation. These standard sensors are easier to configure, maintain, and update compared to specialised 3D cameras, while still achieving accurate identification when combined with other modalities.
Solution Approach 2:
The system replaces complex mechanical 3D camera systems with a combination of standard 2D imaging sensors and computational algorithms. This substitution simplifies the hardware while maintaining or improving identification capability through software-based processing.
3Reliability
If multiple identification systems are used, then identification coverage can be improved, but timing issues arise causing systems to be out of synch
Solution Approach 1:
The system uses feedback mechanisms to continuously monitor and adjust timing across different identification systems. By comparing timestamps from visual, RFID, and sensor systems, the system can detect and correct timing drift, ensuring all systems remain synchronized while maintaining comprehensive identification coverage.
Data Source
Figure 1
Figure 2a
Figure 2b
AI summary
The present invention provides a method for identifying an individual an animal from a plurality of animals, the method comprising the following steps: a) obtaining, via at least one animal recording module at a first time over a first time window, data associated with the plurality of animals, the data comprising a video from a two-dimensional imaging sensor; b) obtaining, via animal identification module at a second time over a second time window, data associated with an identifier of each of a subset of the plurality of animals; wherein the animal recording module is located at a first location and the animal identification module is located at a second location, wherein the first location is spaced apart from the second location; wherein the at least one animal recording module is configured to identify individual animals within the first time window and associate the individual animals with a time stamp corresponding to a point of detection for that animal; wherein the at least one animal identification module is configured to identify individual animals within the second time window and associate the individual animals with a time stamp corresponding to a point of detection for that animal; c) correlating the identified individual animal identified by the at least one animal recording module at the first time window with the animal identified by the at least one animal identification module at the second time window; d) adjusting the width of the second time window relative to the first time window and/or an offset of the second time window with respect to the first time window; repeating steps a) to c) with the adjusted width and/or offset of the time window.