This paper aims to build a pipeline for processing aerial images of precast reinforced concrete elements stored in an outdoor storage area into a map which can be employed to locate concrete elements via object detection. The main focus lies on preprocessing to ease the image stitching process and avoid leveraging positional data. The aerial images are taken by wide angle cameras mounted to a gantry crane. The images are processed by using methods of monocular depth estimation, processed into a point cloud and finally ortho-rectified for stitching. The final result after stitching is a complete map of the storage area. Since the resulting map is to be used for object detection to locate the concrete elements, the images are compiled into a dataset for the detection of concrete elements. The results show that wide angle cameras are ill-suited for a purely image-based approach. The lack of depth estimation models for wide angle cameras hinders last step of orthorectification. Having recognized the importance of positional data, an approach utilizing wide angle cameras augmented by real time kinematics for positional data is suggested.