Tracking of tree logs from a harvesting site to its processing site is a legal requirement for timber-based industries. Wood log identification is an important task in the forestry industry and has traditionally relied on manual inspection by trained experts. However, with the increasing demand for wood products and the need for efficient and accurate identification, there has been a growing interest in developing automated wood log identification systems. In this study, we explored approaches to wood log identification with three objectives: automated log identification, damaged log identification (detection of damaged log), and pith detection (estimation of the pith location). We propose a novel approach for wood log identification using computer vision techniques. Our approach involves capturing images of wood logs using a digital camera, detection of pith location using machine learning model, then applying image processing algorithms to extract relevant features, such as the bark pattern, rings, and use information provided by radial line that extends from the bark to the pith to identify the wood log. In our experiments, we used different image processing techniques to build computer vision models for log identification, trained machine learning and deep learning models to classify the wood log into damaged or not, and deep learning models for estimation of the pith location. The findings of this study indicate that the proposed approach for wood log identification work best with Canny Edge Detection technique, and we can further extend this approach.
Identifer | oai:union.ndltd.org:UPSALLA1/oai:DiVA.org:lnu-124576 |
Date | January 2023 |
Creators | Farooq, Muhammad |
Publisher | Linnéuniversitetet, Institutionen för datavetenskap och medieteknik (DM) |
Source Sets | DiVA Archive at Upsalla University |
Language | English |
Detected Language | English |
Type | Student thesis, info:eu-repo/semantics/bachelorThesis, text |
Format | application/pdf |
Rights | info:eu-repo/semantics/openAccess |
Page generated in 0.0021 seconds