Morphed Picture Recognition using Machine Learning Algorithms
Padmaja Kadiri, Palagati Anusha, Madhav Prabhu, Rolito Asuncion, Voonna Sainath Pavan, Jami Venkata Suman · 2024
With the development of image alteration tools in the digital age, image authenticity has become a major topic. It is becoming more and more difficult to discern between photos that have been altered and those that have not due to the widespread availability of advanced image editing software. A combination of scikit-learn and OpenCV is used to recognize altered photos. In order to detect morphed information, the suggested system analyzes small distortions and changes in images using image processing algorithms. The system extracts pertinent information from the images by using OpenCV for feature extraction and image pre-processing. Then, a machine learning model that can differentiate between real and altered photos is created using scikit-learn. To improve its accuracy and resilience, the model is trained on a dataset that includes both real and altered images. The goal is to advance the area of picture forensics by offering a dependable method for identifying photos that have been altered. To enhance the overall security and reliability of digital content, the technology can be included into pre-existing image verification pipelines in addition to being used independently.