Cork Oak Production Estimation Using a Mask R-CNN

Cork is a versatile natural material. It can be used as an insulator in construction, among many other applications. For good forest management of cork oaks, owners need to calculate the volume periodically. This will allow them choose right time harvest cork. The traditional method laborious and consuming. present work aims automate process calculating trunk area oak from which extracted. Through this calculation, it possible estimate produced before stripping process. A deep neural network, Mask R-CNN, machine learning algorithm are used. dataset images oaks was created, where targets known dimensions were fixed on trunks. R-CNN trained recognize regions, so estimated based target dimensions. Preliminary results show that model presents performance recognition trunks, registering [email protected] 0.96. After obtaining mask results, three models biometric parameters tree. showed support vector average error 8.75%, within margins obtained using methods.