/fcn-potato-scabs-segmentation

Fully Convolutional Network for Semantic Segmentation of Potato Scabs and Mechanical Defect

Primary LanguagePython

fcn-potato-scabs-segmentation

Fully Convolutional Network for Semantic Segmentation of Potato Scabs and Mechanical Defect

Potatoes are an alternative staple and popular source of energy. Not all potatoes in Indonesia have favorable conditions. Diseases in potatoes can significantly damage their value. Common Scab is one of the common diseases that occur in potatoes. An efficient way to detect Common Scab disease in potatoes is by segmenting the potato image. External disease segmentation in potatoes is done using two methods, Deep Learning Fully Convolutional Network and a combination of several digital image processing techniques, including binary thresholding, median blurring, background segmentation, contrast stretching, and color coding.

FCN Result

Manual Segmentation Result (Color Coding)