The diagram is definitely misleading, as is some of the article. There's no high level integration that would allow image recognition, bacteria are AFAIK too s

The diagram is definitely misleading, as is some of the article. There's no high level integration that would allow image recognition, bacteria are AFAIK too simple for that to happen. They may have a cellular automata-like mechanism to detect where the biggest light source is in case there is more than one, or it is diffuse rather than focused. One family of single cell eukaryotes has ocelloids[0] reminiscent of vertebrate and mollusc eyes. Provided nuclear genes form networks conceptually identical to neural nets (i.e. brains), one could imagine that these animals have more elaborate image processing, but I don't know if there's any concrete evidence for it. -- 0. https://en.wikipedia.org/wiki/Ocelloid https://en.wikipedia.org/wiki/Ocelloid