Camouflaged Object Detection (COD) aims to segment objects that are visually integrated into their surroundings. However, due to the high similarity between camouflaged objects and their backgrounds, existing methods often fail to accurately capture subtle boundaries and fine details, leading to suboptimal detection performance. To address this challenge, we propose a novel Boundary-Aware Dual Attention Network, termed BDNet, which aims to enhance boundary perception and improve edge feature learning by jointly leveraging both channel and spatial attention. Specifically, we design an Enhanced Awareness Module that employs the Sobel operator to boost edge sensitivity in shallow feature maps. Then, a Dual Attention Network is introduced, where a scaling factor is employed to balance the contributions of channel and spatial attention, enabling the model to better capture fine-grained boundary details. Moreover, we present an Adaptive Multi-scale Fusion Block that integrates high-level semantic features with low-level detailed features, resulting in more accurate and complete edge representations. Extensive experiments on three benchmark datasets demonstrate that our proposed BDNet achieves highly competitive results compared with 20 state-of-the-art COD methods. The source code is available at https://github.com/zhangz-cyber/BDNet/ .