With the ease availability of digital cameras and access to social media, a huge jump of digital image content can be seen on internet. A digital image can presents a story at a glance and so the forge image may lead to misunderstanding and can create chaos. The copy-move forgery (CMF) is one of the easiest method to forge any digital image due to various tools availability. A robust CMF detection tool can help an individual or organization to verify the authenticity of a digital image. We proposed DeepCMFD method to verify the nature of an image by capturing its low level as well as high-level features with the help of convolutional block and attention block. The proposed method shows the capability of scaling and rotation invariant nature to track the forged regions. Furthermore, the performance of the DeepCMFD method has not impacted due to the post-processing operations performed on the forged images. The experiments have been carried out on CASIA-CMFD and MISD datasets and the obtained results proves the robust nature of DeepCMFD method.