3D Animal Gait Analysis for Reliable Lameness Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing lameness detection systems in animals, particularly in cows, are inaccurate and labor-intensive, often leading to false positives or false negatives due to insufficient image processing and reliance on two-dimensional video analysis, which fails to discriminate between animals and is influenced by low walking velocities.
Innovation Solution
A method using image processing and machine learning to analyze animal locomotion, incorporating velocity as a classifier to improve reliability by focusing on high walking velocities and using 3D cameras to detect abnormalities in the locomotor system, with optional RFID for animal identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If visual observation method is used for lameness detection, then labor intensity is reduced, but detection accuracy deteriorates leading to subjective and late detection
Solution Approach 1:
The patent replaces manual visual observation with an automated image processing system using cameras and machine learning algorithms. The system captures video sequences of animals walking and automatically analyzes gait patterns, substituting human labor with mechanical/optical detection systems to achieve both high efficiency and objective accuracy.
Solution Approach 2:
The patent introduces video recording and image processing algorithms as intermediary elements between the animal's movement and the detection result. Instead of direct human observation, the system uses captured video sequences processed by machine learning models as intermediaries to objectively analyze gait abnormalities and generate detection outcomes.
2Device complexity
If two dimensional video analysis is used, then system complexity is reduced, but discrimination accuracy between animals deteriorates
Solution Approach 1:
The patent transitions from two-dimensional video analysis to three-dimensional spatial analysis by capturing depth information and analyzing spatial relationships between animals. The system processes three-dimensional coordinates of animals in video sequences, enabling accurate discrimination and tracking of multiple animals by utilizing their positions in three-dimensional space rather than just two-dimensional image planes.
3Quantity of substance
If low walking velocity data is included in analysis, then data quantity is increased, but detection reliability deteriorates due to noise from non-locomotor movements
Solution Approach 1:
The patent applies different processing quality standards to different velocity ranges in the data. Instead of uniformly processing all video data, the system identifies and selectively analyzes segments where animals walk at velocities above a predetermined threshold, discarding or giving less weight to low-velocity segments that contain noisy non-locomotor movements. This localized quality approach ensures high detection reliability by focusing on clean, relevant data.
Data Source
Figure 1
Figure 2~3
Figure 4~6
AI summary
The invention is directed at a method of detecting an abnormality in a locomotor system of an animal by means of motion analysis, using a motion analysis system comprising a camera configured for monitoring at least a part of a pathway wherein the animal travels, wherein the camera is operatively connected to a computing device and a data storage of the analysis system, and wherein the computing device comprises a processor and wherein the computing device is communicatively connected to the data storage. The invention is further directed at a motion analysis system.