Computer vision as a method for analyzing journalistic photographic coverage

the 2022 presidential elections in images from the Folha de S. Paulo website

Abstract

The present work aims to rethink methodological approaches to journalistic photographic coverage from the perspective of digital methods (Rogers, 2013). We therefore start from Schneider’s (2015) proposal, which explains the need to rethink photojournalism through narrative valorization and the plurality of images that it contemporary photojournalism raises. This is because, according to Junior (2014), this genre photography has been going through a crisis that directly affects the way photography is done journalism in contemporary times: with newsrooms increasingly emptied, text reporters are forced to take photographs and photographic reporters, to meet a demand for video of new ways of consuming news; journalism loses, when incorporated into the logic of platforms, control over their distribution, monetization and curation, as these pass to obey the logic of algorithms (Van Dijck et al., 2018) and to increasingly consume incidental to its content, as news now competes for public attention with the most different types of publication (Mitchelstein et al., 2016).