Development of a mobile application to autodetect medicinal plants using an artificial intelligence approach

Simple item page
dc.contributor.advisorNgwenya, S.
dc.contributor.advisorSibandze, P.
dc.contributor.authorSimandla, Onke
dc.date.accessioned2026-09-03T07:30:23Z
dc.date.available2026-09-03T07:30:23Z
dc.date.issued2025
dc.descriptionMasters dissertation
dc.description.abstractA mobile application assisting traditional healers in identifying medicinal plants is presented. Python programming language was used to build the mobile application that would auto detect five medicinal plants (Mpepho, Mhlonyane, Hlaka-Hlakana, Mthoba & Cima mlilo) on a Spyder IDE version 5.4.0. The performances of PyTorch and TensorFlow lite Python libraries were tested using the Convolutional Neural Network Deep learning method. Five medicinal plants were part of the data set to be collected using a Samsung cellphone to capture images at different angles. Data sets were collected from their natural habitats at angles 30%, 60%, and 90% per plant species. The best-performing model was selected and presented to the mobile application. Offline capability was programmed into the mobile application, making it possible to operate the mobile application in areas with no network. A machine-learning database used for plant identification was created and stored on the device. On-device storage gave the mobile application the ability to perform plant identifications offline. The offline capability of mobile applications is crucial as the traditional healing community often picks up traditional medicine in jungles where telecommunications networks can be a challenge. Mobile applications achieved an identification result of 98% accuracy, with the 90% angle being the most successful. The most preferred time of the day was the afternoon.
dc.identifier.urihttp://hdl.handle.net/20.500.11837/5091
dc.language.isoen
dc.publisherUniversity of Fort Hare
dc.subjectArtificial Intelligence
dc.subjectMedicinal Plants
dc.subjectMobile Application
dc.titleDevelopment of a mobile application to autodetect medicinal plants using an artificial intelligence approach
dc.typeThesis

Files

Original bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
OSimandla thesis-Final (Examiner comments).pdf
Size:
42.4 MB
Format:
Adobe Portable Document Format
Download
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
1.71 KB
Format:
Item-specific license agreed upon to submission
Description:
Download

Collections

Theses and Dissertations