<![CDATA[Traditional cakes are one of the cultural products that deserve to be treated equally like other cultural assets, including by acknowledging them and preserving their existence. One way is to reintroduce traditional food through digital image processing technology. Taking an image allows us to get the information contained in it. Image segmentation is a technique developed to extract information contained in images which can be used as a reference source. This study proposes the application of traditional cake image segmentation using otsu thresholding method in the CIE LAB color space as extraction of color features. RGB images are used as initial images which are converted into LAB images, each LAB channel then converted into an otsu image. The results of the otsu thresholding image will be compared with the image of ground truth. The results of the study showed that the test values on groundtruth images have different effects on spesific colour of each channel. Of the three channels tested, channel A has the highest accuracy value of 89.65% as well as its specificity and sensitivity values of 87.825 and 95.818%. This indicates that channel A is a channel that can be used on common objects for segmentation with good results than L and B channels.]]>
Copyrights © 2019
