Automated text layout, color and other stylization on an image or video, and the tracking and application of user color preferences

Abstract

Devices and methods for overlaying text content on photos are provided. A digital image is selected. Unique foreground colors from the image are extracted, filtered and color clusters are identified for each foreground color. Color frequency and cluster count within the image are calculated. Text is entered by a user or provided programmatically. Multiple candidate locations for text content are evaluated to identify one or more optimal caption locations on the digital image. One or more suggested text colors are identified for each of the candidate location regions, based on factors including color contrast with background image content at the candidate text location region. The optimal text location regions and colors can be utilized to overlay text content onto the initial digital image to generate an output image. The process can be applied to sampled frames from a video feed. The process can be applied to a mobile device camera application by sampling and compressing frames from an electronic viewfinder video feed, overlaying text content on the preview video feed, capturing a full-resolution image in response to actuation of a camera shutter button, and processing the full-resolution image to yield a final full-resolution captioned output. The result can be transmitted to a social network service or application.