logo

visit!! Assoc. Prof. Dr. Watthanapong Kerdthongmee, Computer Engineering Program Gave interviews and recorded programs on the NBT channel on research development. “Wood core detection system on cross-sectional images of rubberwood logs. using deep neural network techniques”

On February 15, 2023, the NBT television station interviewed and recorded the program. Associate Professor Dr. Watthanapong Kerdthongmee, Lecturer in the Computer Engineering and Artificial Intelligence Program to the results of the research project “Automated system for processing rubber logs into wood planks that increases efficiency using artificial intelligence.”

Assoc. Prof. Dr. Watthanapong Kerdthongmee and Asst. Prof. Korakot Suwannarat, researchers from the Computer Engineering and Artificial Intelligence program. School of Engineering and Technology Walailak University (Mass.) researches and develops a wood core detection system on cross-sectional images of rubberwood logs. Using deep neural network techniques. To connect to the system in the rubber wood processing factory. Helps reduce the amount of rubber wood loss from baking. From the process of cutting a large amount of wood that has a core or filling attached to it. This research has now been published in an international journal in the Scopus Q1 database and is prepared to be further developed into a system for detecting strabismus in humans.

The processing of rubber wood in Thailand today requires human labor called Mr. Ma and Hangma, who control and observe the wood core entering the machine to cut it into wood sheets. Therefore, it requires experience and expertise. At present, there is a shortage of workers with specialized skills in this field. In addition, the characteristics of rubber wood with the wood filling or core are crooked and inconsistent when the horse master and the horse’s tail control the cutting. This results in more wood panels attached to the wood core than usual. When those woods are put into the baking process The core will break out of the wood. This causes the piece of wood to lose value or become unsellable. It must be returned to incineration to rebuild the boiler system. While the quality and value of the wood is quite high. Currently, there is wood damaged from cutting into the core and this type of baking valued at over 500,000 baht/month or over 6 million baht/year.

Rubber wood is the main raw material for producing furniture, parquet, and children’s toys. Including exports in the form of logs and processed wood. Which is the main export product of Thailand that can be exported without quantity limitations continuously. These factors cause rubber wood exports to generate revenue for the country as much as tens of billions of baht. But in the process of sawing logs, there may be mistakes in sawing and attaching the wood filling, which causes the wood to be of poor quality. Associate Professor Dr. Wattanapong Kerdthongmee and the research team therefore had the idea of ​​using artificial intelligence to solve this problem.

 

รศ.วัฒนพงษ์

Assoc. Prof. Dr. Wattanapong Kerdthongmee

Wattanapong Kerdthongmee, head of the research project, revealed that the invention of the wood core detection system on cross-sectional images of rubberwood logs consists of two systems: the log rotation subsystem. and a system for driving logs into the saw system with an AI Camera or artificial intelligence camera that is processed using a deep neural network. Make the system intelligent by creating artificial intelligence. By teaching from more than 400 sample images of rubberwood logs obtained from business establishments, the AI ​​camera captures images of the cross-section of the log, both the head and the tail. As the log continues to rotate to detect in real time This results in wood core detection with an accuracy of up to 100% with an error of no more than 2 millimeters. The saw cuts the wood core area into the thinnest pieces of wood. But it covers the entire wood core in one sheet. Makes it possible to get the most amount of good quality wood. Greatly reduces the rate of wood loss. It is also fully automated, reducing manual labor to just one person.

นวัตกรรมตรวจจับไม้

“This camera is considered a novelty in the rubber wood processing industry and Thai agricultural industry. Because it uses artificial intelligence The distinctive feature of this system is that it cannot be replaced by foreign software. This is because rubber wood has a unique characteristic in that the core deviates from the center of the log, is crooked, or is not truly cylindrical. and contamination from latex, saw marks, and fungus on the cutting surface. It is considered a problem of rubber wood processing that is a specific problem in Thailand. Foreign machines designed to cut oak or pine with straight cores cannot be used instead,” Assoc. Prof. Dr. Wattanapong said.

นวัตกรรมตรวจจับไม้
Facebook Comments Box