Abstract: An image processing method generates a training dataset for training a machine learning model to predict illumination conditions for different positions relative to a scene, the training dataset including training images and reference data.

Abstract: Synchronization of two video streams that have been processed in different ways is achieved by generation of logical maps representative of characteristics, such as differences, between sample values and their spatial neighbors in a current stream and in a reference stream. For samples in a current stream and co-located samples in the reference stream, logical maps are generated. Those frames in each stream that have the best fit regarding equal logical map values are aligned to synchronize the streams.