Education is a fundamental right that forms the backbone of every child and is, therefore, crucial for it to be available to everyone without any discrimination. Our primary aim with this dissertation is to make general school textbooks used by regular children available for visually impaired students so that they are not deprived of the content and knowledge learned by others. We built an image caption generating model using deep learning which is used to generate text from images of school textbooks, storybooks, and other picture books when passed into the model. We can then use the text generated and convert it into an audio format in different languages which can then help visually impaired students use general textbooks in an audio format. We used the pre-trained image captioning model VGG16 to generate captions. We built our dataset of animated textbook images from scratch and trained our model using the same. We successfully converted the text obtained into an audio format of different languages like English, Hindi, Tamil, Malayalam, etc.