WOOD DEFECT DETECTION UNDER REGION OF INTEREST TARGET DETECTION ALGORITHM

Abstract

Good quality wood will provide maximum product results for manufacturers of wooden products such as furniture. Determining the quality of wood to avoid wood defects has now become a special need for wood product manufacturers and craftsmen. This research aims to detect wood defect objects in digital image results to determine wood quality. The detection process will involve the performance of Digital Image Processing (DIP) using the Region Of Interest (ROI) method. The performance of the ROI method can describe image objects based on the object area which can be determined using a masking process on the specified area values. The research dataset uses photographic images of wooden objects which have a pixel resolution of 1024 x 1024, a total of 207 images. Based on the tests that have been carried out, the output from the detection process presents a fairly good image of wood defect objects. The output images of wood defective objects can depict the image of wood defective objects precisely and accurately. The performance of the ROI method in detecting wood defective objects has worked optimally so that wood quality can be determined based on the image output. The results of the detection image output in this research can later be used as a form of recommendation for related parties, especially for wood producers and craftsmen in determining quality wood to increase productivity