Automated Livestock Weight Estimation Using 2D Camera Imaging

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for accurately estimating the weight of livestock animals are impractical and stressful for the animals, leading to reduced yield and variability in accuracy, with existing technologies like 3D imaging and visual analysis showing limitations in precision and reliability.

Innovation Solution

A device using embedded processing means and a two-dimensional camera to calculate parameters such as size, weight, and volume of livestock animals, employing an a priori model and heuristic calculations, with optional remote processing and stereoscopic vision, along with segmentation techniques for image analysis.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If scales are used to accurately weight animals, then measurement precision is improved, but ease of operation deteriorates due to logistical impracticality and animal stress

Engineering Contradiction:
Improveweight estimation accuracyVSAvoidlogistical practicality
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The patent replaces mechanical weighing scales with an automated image processing system using two-dimensional cameras and computational algorithms to estimate animal weight, thereby eliminating the need for physical scale placement and animal handling while maintaining measurement accuracy

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system creates a digital copy of the animal through two-dimensional imaging and uses this copy for weight estimation calculations, eliminating the need for direct physical measurement with scales while preserving measurement accuracy

Inventive Principle:
Principle #26Copying

2Ease of operation

If visual evaluation by producers is used to estimate animal weight, then ease of operation is improved, but measurement precision deteriorates with root mean squared error of about 6.4 Kg

Engineering Contradiction:
Improveweight estimation simplicityVSAvoidweight estimation accuracy
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent replaces subjective human visual evaluation with an automated computer-based image processing system that objectively analyzes animal images and calculates weight estimates, eliminating human error while maintaining operational simplicity

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables producers to independently obtain accurate weight estimates through automated image analysis without requiring specialized training or equipment, allowing them to self-assess animal weight with scientific precision

Inventive Principle:
Principle #25Self-service

3Measurement precision

If 3D imaging is used to estimate animal weight, then measurement precision is improved, but device complexity increases

Engineering Contradiction:
Improveweight estimation accuracyVSAvoidimaging system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent extracts only the essential two-dimensional projection information needed for weight estimation, eliminating the complexity of full three-dimensional imaging systems while maintaining sufficient measurement precision for practical applications

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The system uses standard two-dimensional cameras instead of expensive 3D imaging equipment, accepting that the images are single-use snapshots rather than continuous 3D models, thereby reducing device complexity and cost while achieving practical measurement accuracy

Inventive Principle:
Principle #27Cheap short-living objects (Disposable)

Data Source

PatentEP3158289B1Method and device for automated parameters calculation of an object
Publication Date: 2024.05.01 TOUCHLESS ANIMAL METRICS SL
  • EP3158289B1 patent drawingFigure 1~2
  • EP3158289B1 patent drawingFigure 3~4
  • EP3158289B1 patent drawingFigure 5a

AI summary

the method comprising: acquiring, by a two-dimensional camera (7), in a scene, a two dimensional image of at least one object (200); identifying the object (200) within the acquired two dimensional image; calculating, by a first means, the size of a pixel of the object (200) in the acquired and segmented two dimensional image taking into account the distance between the object (200) and the two-dimensional camera (7); and calculating, by a second means, several parameters including at least the size, dimensions, body part dimensions, body features, weight and/or volume of the object (200) by using said calculated size of the pixel and an a priori model of the object (200), wherein said a priori model includes information linking different parts, contours or shapes representative of several objects (200), previously acquired with a two- dimensional camera, with several parameters said several objects (200).