Redescribing Images: A Novel Approach Using Vision Transformer and Image Enhancement Techniques

An image caption is a brief description or explanation that goes along with an image. Image captions help to improve theway visual information is conveyed, making it easier to understand and comprehend. Traditional machine learning models of feature extraction and deep learning models of the encoder-decoder model have some drawbacks of limited scalability and produce a generic description of the images. Getting a description through a vision transformer is a challenging task. To enhance the quality of the image and the semantic description of the image, apply one of the image enhancement techniques, Morphological Open. This paper discusses redescribing an image by applying image enhancement techniques through a vision transformer, which is a novel approach to extracting images.