Camouflage Object Detection (COD) aims to identify objects that perfectly blend with the surrounding environment from complex backgrounds. Identifying disguised objects that are visually highly integrated with the background is a highly challenging task. To address this challenge, we propose a COD framework called Iterative Network Based on Separable Attention (ISANet), which refines low resolution features through iterative feedback using high-resolution features. Our goal is to extract high-resolution texture details to avoid feature degradation that leads to blurred edges and boundaries. To handle background clutter and occlusion artifacts, we propose the foreground and background separation attention mechanism which separates foreground and background through a multi-stage iterative method to reinforce position information, thereby improving detection accuracy and enhances the ability of distinguishing objects from complex backgrounds. Extensive experiments conducted on three challenging datasets demonstrate that our ISANet achieves state-of-the-art COD performance.