Data Augmentation Method of Object Detection for Deep Learning in Maritime Image
Hyeon-Cheol Shin, Kwang-Il Lee, Chang-Eun Lee
Abstract
Maritime object detection is an essential element for the situational awareness in autonomous ships. Recently, as the deep learning technology evolves, the attempt that the ship recognize the surrounding environment by using deep learning technology is gradually increasing. Deep learning methods, however, require a lot of data, but lack a publicly available dataset for object detection in the maritime domain. In this paper, we proposed a data augmentation method that can automatically extend the object detection dataset in maritime image. Extract the mask of the foreground object and combine it with the new background to automatically generate the location information and data of the object. Through the proposed method, we can learn high quality data by configuring various limited data features.