Chertify: Wood Identification-Based Mobile Cross-platform by Deep Learning Technique

Simple item page
dc.contributor.authorWongpoo, Teerasak
dc.contributor.authorSriwan, Wannamongkol
dc.contributor.authorTitijaroonroj, Taravichet
dc.contributor.authorJamsri, Pornsuree
dc.date.accessioned2026-08-06T10:35:43Z
dc.date.available2026-08-06T10:35:43Z
dc.date.issued2022-01-01
dc.description.abstractThailand’s economic trees are counted as one of its most valuable domestic assets and well known internationally as a high quality natural wood resource. However, there is a need for basic wood identification whether or not for a required certificate by individuals, entrepreneurs, and organizations. Currently, the wood identification process is manually accomplished only by an expert at the Forest Research and Development Office, the Royal Thai Forest Department. This is a time consuming complex process for two reasons–required experience and limited experts. Given the complexity of wood identification, a new approach is offered, namely, to identify different types of wood with an image from a smartphone. The researcher initially proposes a mobile application, “Chertify”, that has five features (login, wood check, wood check history, manual, and wood knowledge). This app can serve both iOS and Android platforms and targets the general user. Chertify aims to simplify identification of an economic wood type by combining deep learning technology with an actual wood image on a Smartphone. The selected deep learning algorithm will be applied to 258 trained group images of seven wood types based on highest accuracy and lowest standard deviation. Chertify relies on a handcrafted method (HOG and SVM) and a learning-based method (Alexnet) with accuracy of 69.4% and 84.73% and SD at 5.37% and 3.07% values, respectively.
dc.identifier.citationLecture Notes in Networks and Systems, 453 LNNS, 77-87, 2022
dc.identifier.doi10.1007/978-3-030-99948-3_8
dc.identifier.issn23673370
dc.identifier.other2-s2.0-85128540416
dc.identifier.urihttps://dspace.kmitl.ac.th/handle/123456789/12879
dc.sourceLecture Notes in Networks and Systems
dc.subjectDeep learning
dc.subjectMachine learning
dc.subjectMobile application
dc.subjectWood identification
dc.titleChertify: Wood Identification-Based Mobile Cross-platform by Deep Learning Technique
dc.typeConference Paper

Files

Collections

Research