A camouflaged object detection method based on Convnextv2 and texture-edge guidance was proposed in order to address the issue of insufficient expression and processing of edge features of targets and unique texture feature information in corresponding scenarios in camouflaged object detection. The texture encoding module was used to extract texture features from input images, which were fused with the edge features extracted by the backbone network to generate texture-edge features of the images. The texture-edge features were integrated into the backbone features to locate the true position of the target through the designed texture-edge guided attention module. A feature fusion module was employed for multi-level feature fusion, and a multi-level supervision approach was adopted to design the overall loss function. Experiments on three public datasets (CAMO, COD10K, NC4K) and the camouflage mixed dataset MICAI_TE showed that the algorithm achieved optimal comprehensive performance.