Automatic Image Alignment and Fusion in a Digital Photomontage

Abstract

A composite of several pictures is what we call a Photomontage or a collage. Having a large collection of photographs from an event, wedding, or trip is common these days. It is tedious to manually select meaningful images from the large collection of pictures to create a summary or story-telling montage. With the rapid development of computer vision algorithms, the modern digital world operates on automated and intelligent vision systems. Regular users highly demand these intelligent, automatic, and user-friendly systems. Our focus is to create an intelligent and user-friendly system that requires little to no manual work to create a digital photomontage. As a result, this paper describes the procedure of an automated digital photomontage web app. This system aims to construct an eye-pleasing image from a collection of input images without any effort. The system constructs a montage, also known as collage, of given images using several computer vision algorithms to identify each image’s right placement. The algorithms involve saliency detection, face detection, smooth image-to-image transition, auto-orientation, and image-borders blending. The web app is available at https://jsvirk47.pythonanywhere.com and code is available at https://tinyurl.com/h4ccehrx. The proposed system is easy to use and provides high-quality results. This system is ideal for regular users to process images to share with their friends and family or social media profiles.