Pneumonia is a disease characterized by lung infection, typically instigated by fungi, bacteria, or viruses. The infection affects the lungs air sacs and the oxygen cannot enter into the blood cells. Pneumonia majorly affects children, causing severe health problems and it leads to death. To detect pneumonia there are many algorithms, but our approach is to provide the efficient algorithm to detect the pneumonia at the early stage. Pneumonia is detected by giving the chest X-rays images as input and the output is obtained as whether the pneumonia is present or not. In this paper, provided an effective algorithm in deep learning called Extended VGG19 (Extended Visual Geometry Group) which comes under the category of CNN (Convolutional Neural Network). In this approach, the VGG19 model was extended with three additional convolutional layers to quickly detect the pneumonia. The Extended VGG19 model was trained on a diverse dataset of chest X-ray images, including both normal and pneumonia-affected cases, using transfer learning. It was observed that, in comparison to the standard VGG19 architecture, significant improvement in pneumonia detection 99.12% accuracy was achieved by the Extended VGG19 model.