Implementation of Plant Image Recognition Machine Learning API
Honors Capstone Project
1
Advisor(s)
Dr. Ahmed Oun
Confirmation
1
Document Type
Paper
Location
McIntosh Activities Room
Start Date
16-4-2024 2:15 PM
End Date
16-4-2024 5:00 PM
Abstract
One common issue novice plant owners have is not knowing their plant’s common name or genus. This hinders plant owners from educating themselves to better care for their plants as they are unable to easily search for the relevant data and often leads to experimentation at the cost of the owner’s time and money. This presentation presents an implemented API that facilitates plant identification and information retrieval within a mobile application. Utilizing Machine Learning (ML) image recognition technology, users can upload plant images using an app for quick analysis. Once the user uploads an image, it is processed through a ML algorithm by Plant.id, an external plant identification service. Upon successful identification, relevant information such as common name, species, genus, and description is retrieved, formatted, and then displayed to the user. The application incorporates two databases: "Leaflife" and "Plant.id." If the plant is not recognized by the capstone team’s database, "Leaflife," the system refers to the previously received "Plant.id” database response for the pertinent details. The user interface dynamically updates to reflect the identified plant's information, allowing for a seamless user experience. To optimize performance and responsiveness, asynchronous programming was employed throughout the implementation to ensure smooth handling of image recognition requests, database queries, and UI updates. Additionally, the system leverages post-frame callbacks to synchronize UI updates with the completion of asynchronous tasks, maintaining consistency and usability. Overall, the implementation streamlines the user’s experience, by providing an intuitive and efficient tool for learning about the botanical world.
Recommended Citation
Flores, David, "Implementation of Plant Image Recognition Machine Learning API" (2024). ONU Student Research Colloquium. 25.
https://digitalcommons.onu.edu/student_research_colloquium/2024/papers/25
Level of Access
Restricted to ONU Community
Restricted
Available to ONU community via local IP address and ONU login.
Implementation of Plant Image Recognition Machine Learning API
McIntosh Activities Room
One common issue novice plant owners have is not knowing their plant’s common name or genus. This hinders plant owners from educating themselves to better care for their plants as they are unable to easily search for the relevant data and often leads to experimentation at the cost of the owner’s time and money. This presentation presents an implemented API that facilitates plant identification and information retrieval within a mobile application. Utilizing Machine Learning (ML) image recognition technology, users can upload plant images using an app for quick analysis. Once the user uploads an image, it is processed through a ML algorithm by Plant.id, an external plant identification service. Upon successful identification, relevant information such as common name, species, genus, and description is retrieved, formatted, and then displayed to the user. The application incorporates two databases: "Leaflife" and "Plant.id." If the plant is not recognized by the capstone team’s database, "Leaflife," the system refers to the previously received "Plant.id” database response for the pertinent details. The user interface dynamically updates to reflect the identified plant's information, allowing for a seamless user experience. To optimize performance and responsiveness, asynchronous programming was employed throughout the implementation to ensure smooth handling of image recognition requests, database queries, and UI updates. Additionally, the system leverages post-frame callbacks to synchronize UI updates with the completion of asynchronous tasks, maintaining consistency and usability. Overall, the implementation streamlines the user’s experience, by providing an intuitive and efficient tool for learning about the botanical world.