<p>Finding and segmenting objects that are visually heavily integrated with the background is the aim of camouflaged object detection. Owing to the significant resemblance between camouflaged objects and their surroundings, there are still issues of imbalance between global and local information, as well as blurry or unclear object boundaries. Therefore, we propose a novel integrated feature fusion and boundary optimization network, named IFBONet. Specifically, we first employ the Swin Transformer as the backbone to extract rich global context information. Secondly, we propose a focus intersection decoder (FID), which aims to approximate the initial position of the camouflaged object by leveraging dense intersection strategies. Building on this foundation, a feature hybrid interaction module (FHIM) is introduced, which effectively aggregates multi-scale features using a hybrid interaction tactic, ensuring thorough information exchange among different features. Finally, the boundary refinement module (BRM) is used to further obtain rich boundary information and perform noise filtering through the edges to yield the final refined prediction map. Extensive experimental findings on four benchmark datasets confirm the efficacy of our proposed IFBONet network, which achieves significant performance improvements compared to 15 other state-of-the-art methods.</p>