GRAPHIS
Visualise, Draw, Annotate, and Save Image Regions in Graffiti Photos
Abstract
A digital photo file contains the image pixel values along with associated photo metadata. Storing those metadata is enabled by various standards. For instance, the Exif standard enables the recording of technical photo metadata like the camera’s serial number and focal length, while the IPTC Photo Metadata Standard is the widely accepted norm for storing copyright and descriptive information in images (from unedited photos to AI-generated pictures). Since its 2019.1 version, the IPTC Photo Metadata Standard has facilitated the creation of image regions: groupings of image pixels—defined by a circle, rectangle, or any other polygonal shape—which can be annotated with region-specific metadata. Given the potential of image regions for graffiti photo annotation, the open-source and freely available software GRAPHIS was developed within the academic graffiti project INDIGO. GRAPHIS (Generate Regions and Annotations for PHotos using the IPTC Standard) allows users to generate and visualise image regions, annotate them with graffiti descriptions or transcriptions, and save them as metadata within the image. To adhere to the IPTC Photo Metadata Standard at every stage, project INDIGO also created a dedicated controlled vocabulary to contain all relevant concepts that can be used to define each image region’s role and content type. This paper starts with a general overview of metadata concepts, followed by a more in-depth look at Exif and IPTC photo metadata. After describing the IPTC Image Region property, the text details the workings of GRAPHIS and the controlled vocabulary development. An overview of use cases and potential software improvements conclude the text.