On the Integration of Synthetic Objects with Real-World Scenes

Abstract

This paper presents a method for integrating synthetic objects in real-world photographs. We take a set of photographs of a real scene and build a simple image-based model. We use high dynamic range images to build an accurate representation of the lighting in the scene. Then we insert a synthetic object into the model and compute its illumination and shading using the lighting information. Illumination changes produced by the synthetic object are also applied to the background plane in the real-world photograph. We show how easy it is to achieve photorealistic results without specialized hardware. Our approach takes advantage of techniques like automatic camera calibration, high-dynamic range image capture and image-based lighting. We show preliminary results obtained with our application. We also present two improvements that we are currently studying. They are aimed at improving the quality of the resulting images and decreasing the computational resources needed by the process.