The Digital Rock Workflow leverages cutting-edge imaging technologies and state-of-the-art image processing algorithms to create digital representations of reservoir rocks. These digital models serve as the foundation for conducting physics simulations, enabling the calculation of various petrophysical properties, such as porosity and permeability. 

The effectiveness of the entire Digital Rock workflow is intricately tied to the resolution at which images are acquired, a parameter currently constrained by the limitations of micro-CT scanning technology. Moreover, capturing high-resolution, large-field-of-view images for heterogeneous rock samples with a substantial Representative Elementary Volume (REV) can be a complex and time-consuming endeavor. As a result, there is a pronounced emphasis on the development of computational techniques aimed at digitally enhancing the resolution of these rock volumes.  

Visual tracking is one of the most challenging problems in computer vision. Given the object of interest in the first frame, the role of the tracker is to follow the trajectory of the target. However, there are several challenges such as scale variation, illumination changes, blurring effects and occlusion which are yet to be tackled with sufficient degree of realism. In this project we addressed scale variation by learning a scale parameter which resizes the bounding box (bbox) prediction given by tracker.