Tree Sense Imaging - Web Application Using Deep Learning

Abstract

This is the paper concerning TreeSense Imaging. It is a system that can be applied on the web where one can find and count trees automatically. The model adopted in the system was modified with ONNX technology. This is to say that individuals are able to view trees on their web browser. They are not required to have computers and servers in use. The TreeSense Imaging consists of various components.

• The home page which is the initial one.
• A place where you can post the pictures of trees.
• A page indicating what the system thinks it discovered.
• A history of what the system has accomplished in the past.
• There is a section about information on TreeSense Imaging.

TreeSense Imaging is quite useful to individuals who work with tree and to those who conduct research on trees. It assists them in carrying out the job to the completion. TreeSense Imaging is an application, used by forestry professionals and scientists that map with TreeSense Imaging. This should be more flexible to deploy than the traditional server-based systems because it has better accessibility, faster processing, and it is more flexible. On several datasets of forest images, the accuracy of the testing was recorded to be high exceeding 92 in the detection and counting of trees.