Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Development of Equine Gait Recognition Algorithm

Maga, Mikael LU and Björnsdotter, Sigrid LU (2017) BMEM01 20171
Department of Biomedical Engineering
Abstract
Horseback riding is a sport enjoyed by people around the world. Many riders are interested in knowing exactly how much they have exercised their horse and how much time that have been spent in different gaits. The goal of this master's thesis was to develop an equine gait recognition algorithm. Triaxial accelerometer and gyroscope signals were collected during different riding sessions by using smartphones. Features, used in previous activity recognition works, were implemented and calculated for all sensor signals. Different methods to select important features were used and the feature sets were then evaluated. In the work four classifiers were implemented and evaluated.

The work resulted in an equine gait recognition algorithm based... (More)
Horseback riding is a sport enjoyed by people around the world. Many riders are interested in knowing exactly how much they have exercised their horse and how much time that have been spent in different gaits. The goal of this master's thesis was to develop an equine gait recognition algorithm. Triaxial accelerometer and gyroscope signals were collected during different riding sessions by using smartphones. Features, used in previous activity recognition works, were implemented and calculated for all sensor signals. Different methods to select important features were used and the feature sets were then evaluated. In the work four classifiers were implemented and evaluated.

The work resulted in an equine gait recognition algorithm based on signals collected at the saddle-girth. The developed algorithm used a window length of 128 samples and windows with 50 % overlap. A feature set was chosen by the use of sequential forward feature selection. Five features were included in the final algorithm and two classifiers using two respectively three of the features. The first classifier separated stand from the gaits by using the features root mean square for the magnitude of the gyroscope signal and energy of the x-axis accelerometer signal. The second classifier classified gaits as either walk, trot or canter using the wavelet based feature energy distribution ratio of the z-axis accelerometer signal, dominant frequency of z-axis of the gyroscope signal and skewness of the accelerometer z-axis. The classifiers used in both classification steps were KNN with K = 3.

The algorithm performed well on a collected test set including two riding sessions. It should be noted that the same phone was used to collect both training and testing data. The performance of the developed algorithm was benchmarked against the smartphone application Equilab. The performance of both algorithms was similar. The developed equine gait recognition algorithm had a 94.1 % and 97.4 % accuracy on the two different test sessions.

Further development of the algorithm will be needed to include other terrains and a larger variety of horses and riders. (Less)
Popular Abstract (Swedish)
Håll koll på din ridning

Ryttare har visat stort intresse för en produkt som kan urskilja hur mycket de verkligen motionerat sin häst. I detta arbete har det därför utvecklats en algoritm som kan användas för att urskilja hästens gångarter. Detta har gjorts genom att använda sensorer i en vanlig smartphone.

Ridning är en populär sport som utövas av många världen över. Ryttare lägger ner otaliga timmar på sina hästar och vill att de ska må så bra som möjligt. För att hästar ska må bra krävs att de får mycket motion och att motionen är varierande. Ryttare har visat stort intresse för en produkt där de kan följa upp sin träning och se hur mycket de tränat sin häst och hur mycket tid som tillbringats i de olika gångarterna. Genom... (More)
Håll koll på din ridning

Ryttare har visat stort intresse för en produkt som kan urskilja hur mycket de verkligen motionerat sin häst. I detta arbete har det därför utvecklats en algoritm som kan användas för att urskilja hästens gångarter. Detta har gjorts genom att använda sensorer i en vanlig smartphone.

Ridning är en populär sport som utövas av många världen över. Ryttare lägger ner otaliga timmar på sina hästar och vill att de ska må så bra som möjligt. För att hästar ska må bra krävs att de får mycket motion och att motionen är varierande. Ryttare har visat stort intresse för en produkt där de kan följa upp sin träning och se hur mycket de tränat sin häst och hur mycket tid som tillbringats i de olika gångarterna. Genom samtal med ryttare har det framkommit att upplevelsen av ett ridpass inte alltid stämmer överens med verkligheten och att det skulle vara bra att få svart på vitt hur ett träningspass verkligen sett ut. Detta skulle kunna hjälpa ryttare att att utveckla sin träning och anpassa den efter sin häst. Det finns redan en del produkter ute på marknaden som kan användas för analys av ett ridpass. Det har dock framkommit att dessa produkter haft problem med gångartsigenkänning.

Med denna bakgrund utvecklades en egen algoritm som kan användas för att detektera och urskilja hästens gångarter, skritt, trav och galopp, samt om hästen står stilla. Algoritmen är utvecklad så att den kan säga hur mycket tid som har tillbringats i de olika gångarterna och vid vilka tidpunkter det ridits i de olika gångarterna. Gångartsigenkänningen görs genom att analysera signaler som spelats in från sensorer som finns i en vanlig smartphone. De sensorer som använts är accelerometern och gyroskopet. Accelerometern mäter telefonens acceleration medan gyroskopet mäter rotationen. När algoritmen testades visade den sig fungera mycket bra och kunde urskilja de olika gångarterna utan problem.

Denna algoritm är en bra start inför utvecklingen av en produkt, till exempel en smartphone applikation, som skulle kunna innehålla mer avancerade funktioner. Produkten skulle kunna utvecklas till att innehålla ett mått på ridningens kvalitet eller kunna detektera om hästen är halt.

För att vidareutveckla algoritmen behövs mer data spelas in. Denna data bör bland annat innehålla mer variation, från olika hästar och olika underlag. Detta kommer göra att algoritmen kan fungera väl får många olika typer av hästar och på många olika underlag. (Less)
Please use this url to cite or link to this publication:
author
Maga, Mikael LU and Björnsdotter, Sigrid LU
supervisor
organization
course
BMEM01 20171
year
type
H2 - Master's Degree (Two Years)
subject
keywords
Activity Recognition - Gait Analysis - Horse Back Riding - Accelerometer - Gyroscope
language
English
additional info
2017-13
id
8918412
date added to LUP
2017-06-27 09:19:49
date last changed
2017-06-27 10:03:57
@misc{8918412,
  abstract     = {{Horseback riding is a sport enjoyed by people around the world. Many riders are interested in knowing exactly how much they have exercised their horse and how much time that have been spent in different gaits. The goal of this master's thesis was to develop an equine gait recognition algorithm. Triaxial accelerometer and gyroscope signals were collected during different riding sessions by using smartphones. Features, used in previous activity recognition works, were implemented and calculated for all sensor signals. Different methods to select important features were used and the feature sets were then evaluated. In the work four classifiers were implemented and evaluated.

The work resulted in an equine gait recognition algorithm based on signals collected at the saddle-girth. The developed algorithm used a window length of 128 samples and windows with 50 % overlap. A feature set was chosen by the use of sequential forward feature selection. Five features were included in the final algorithm and two classifiers using two respectively three of the features. The first classifier separated stand from the gaits by using the features root mean square for the magnitude of the gyroscope signal and energy of the x-axis accelerometer signal. The second classifier classified gaits as either walk, trot or canter using the wavelet based feature energy distribution ratio of the z-axis accelerometer signal, dominant frequency of z-axis of the gyroscope signal and skewness of the accelerometer z-axis. The classifiers used in both classification steps were KNN with K = 3.

The algorithm performed well on a collected test set including two riding sessions. It should be noted that the same phone was used to collect both training and testing data. The performance of the developed algorithm was benchmarked against the smartphone application Equilab. The performance of both algorithms was similar. The developed equine gait recognition algorithm had a 94.1 % and 97.4 % accuracy on the two different test sessions. 

Further development of the algorithm will be needed to include other terrains and a larger variety of horses and riders.}},
  author       = {{Maga, Mikael and Björnsdotter, Sigrid}},
  language     = {{eng}},
  note         = {{Student Paper}},
  title        = {{Development of Equine Gait Recognition Algorithm}},
  year         = {{2017}},
}