Deep morphological networks.

Gianni Franchi, Amin Fehri, Angela Yao

Abstract: Highlights • A novel morphological deep learning framework with learned mathematical morphology operators. • We are the first to attempt to learn the weights of non-approximated mathematical morphology operators end-to-end in deep learning frameworks. • A replacement for the standard max pooling in convolutional neural networks with a learned morphological pooling that proves to be experimentally beneficial. • We propose a mixed morphological and convolutional neural network that performs edge detection with results competitive with state-of-the-art. In addition, this network is trained from scratch, on the contrary to state-of-the-art that make use of pretrained weights. • We propose a fully morphological neural network for image denoising that present better performance than similar fully convolutional neural network for this task. Abstract Mathematical morphology provides powerful nonlinear operators for a variety of image processing tasks such as filtering, segmentation, and edge detection. In this paper, we propose a way to use these nonlinear operators in an end-to-end deep learning framework and illustrate them on different applications. We demonstrate on various examples that new layers making use of the morphological non-linearities are complementary to convolution layers. These new layers can be used to integrate the non-linear operations and pooling into a joint operation. We finally enhance results obtained in boundary detection using this new family of layers with just 0.01% of the parameters of competing state-of-the-art methods. Previous article in issue Next article in issue