Images are frequently manipulated for various purposes, often serving the interests of specific parties. Given that images are commonly regarded as evidence of reality, their manipulation can significantly contribute to the spread of fake news or misleading information. Detecting such image falsifications necessitates access to extensive image data and the development of models capable of scrutinizing each pixel within an image. Furthermore, ensuring efficiency and flexibility in data training is vital to support practical applications. Big data and deep learning concepts, particularly the Convolutional Neural Network (CNN) architecture utilizing Error Level Analysis (ELA), have proven highly effective, achieving a forgery detection rate of 91.83% with convergence in just 9 epochs.