End-to-End Trainable Deep Active Contour Models for Automated Image Segmentation: Delineating Buildings in Aerial Imagery

TL;DR

This paper introduces Trainable Deep Active Contours (TDACs), a novel end-to-end deep learning framework combining CNNs and active contour models for precise, automatic building segmentation in aerial imagery, outperforming existing methods.

Contribution

The paper presents a fully differentiable, end-to-end trainable architecture that integrates CNNs with active contour models for improved building segmentation accuracy.

Findings

TDAC achieves state-of-the-art performance on aerial building datasets.

The model provides fast and fully automatic delineation of multiple buildings.

TDAC outperforms existing deep learning approaches in accuracy.

Abstract

The automated segmentation of buildings in remote sensing imagery is a challenging task that requires the accurate delineation of multiple building instances over typically large image areas. Manual methods are often laborious and current deep-learning-based approaches fail to delineate all building instances and do so with adequate accuracy. As a solution, we present Trainable Deep Active Contours (TDACs), an automatic image segmentation framework that intimately unites Convolutional Neural Networks (CNNs) and Active Contour Models (ACMs). The Eulerian energy functional of the ACM component includes per-pixel parameter maps that are predicted by the backbone CNN, which also initializes the ACM. Importantly, both the ACM and CNN components are fully implemented in TensorFlow and the entire TDAC architecture is end-to-end automatically differentiable and backpropagation trainable without user intervention. TDAC yields fast, accurate, and fully automatic simultaneous delineation of arbitrarily many buildings in the image. We validate the model on two publicly available aerial image datasets for building segmentation, and our results demonstrate that TDAC establishes a new state-of-the-art performance.