Formal semantics of natural language perceives sentences as logical expressions. Full paragraphs and even stories of natural language texts are researched and formalized using discourse analysis (Discourse Representation Theory is one example). My question is - is there research trend that applied the notion "discourse" to images, sounds and even animation? Is there such a notion as "visual discourse"?
Google gives very few and older research papers, so - maybe the field exists, but it uses different terms and Google can not relate those terms to my keyword "visual discourse".
Basically - there are visual grammars and other pattern matching methods that can discover objects in the picture and relate them. But one should be able to read whole store from the picture (musical piece, multimedia content) and I imagine that such reading can be researched by multimedia discourse analysis. But there is no work under such terms. How it is done and named in reality?