This paper presents an innovative mobile device-based system for digitizing and automatic alignment of Optical Mark Recognition sheets. The system uses image processing techniques, including grayscale conversion, contrast enhancement, and edge detection, to ensure high-quality alignment of captured images. By employing template-based automatic alignment and perspective warping, the system compensates for distortions such as tilts and rotations, enhancing the accuracy of subsequent analysis. The system was tested in real-world conditions using live image capture, demonstrating its ability to process diverse, unlabeled data in real time without relying on pre-existing datasets or prior training. Experimental results validate the system’s effectiveness in real-time auto-alignment, compensating for rotational offsets of up to 12 \(^{\circ }\) with an average execution time of 1.24 s, this, even under the limited processing capabilities of mobile devices.