Mohajerani Sorour, Saeedi Parvaneh
IEEE Trans Image Process. 2019 Mar 11. doi: 10.1109/TIP.2019.2904267.
Automatic identification of shadow regions in an image is a basic and yet very important task in many computer vision applications such as object detection, target tracking, and visual data analysis. Although shadow detection is a well-studied topic, current methods for identification of shadow are not as accurate as required. In this work, we propose a deep-learning method for shadow detection at a pixel-level that is suitable for single RGB images. The proposed CNN-based method benefits from a novel architecture through which global and local shadow attributes are identified using a new and efficient mapping scheme in the skip connection. It extracts and preserves shadow context in multiple layers and utilizes them gradually in multiple blocks to generate final shadow masks. The training phase of the network is simple and can be directly and easily adapted for other image segmentation tasks. The performance of the proposed system is evaluated on three publicly available datasets sbudataset,stcgan,ucf, where it outperforms the state-of-the-art Balanced Error Rates (BER) by 3%, 6.2%, and 11.4%.
在许多计算机视觉应用中，如图像目标检测、目标跟踪和视觉数据分析，自动识别图像中的阴影区域是一项基础且非常重要的任务。尽管阴影检测是一个研究充分的课题，但当前的阴影识别方法仍未达到所需的精度。在这项工作中，我们提出了一种适用于单幅RGB图像的像素级深度学习阴影检测方法。所提出的基于卷积神经网络（CNN）的方法受益于一种新颖的架构，通过该架构，在跳跃连接中使用一种新的高效映射方案来识别全局和局部阴影属性。它在多个层中提取并保留阴影上下文，并在多个模块中逐步利用这些上下文来生成最终的阴影掩码。网络的训练阶段简单，并且可以直接且轻松地适用于其他图像分割任务。所提出系统的性能在三个公开可用的数据集sbudataset、stcgan、ucf上进行了评估，在这些数据集上，它的平衡错误率（BER）比当前最先进的方法分别高出3%、6.2%和11.4%。
