3 ms·
> The police may not have a good way to evaluate the vastly superior dynamic range of human vision compared to the camera. The solution is simple and even poin
by imh 9y ago
> The police may not have a good way to evaluate the vastly superior dynamic range of human vision compared to the camera.
The solution is simple and even pointed out later in the article:
> Note that the streetlamps are actually not that far from her crossing point, so I think she should have been reasonably illuminated even for non-HDR cameras or the human eye, but I would need to go to the site to make a full determination of that.
Seems like the question of whether a driver would have seen the pedestrian is simply solved by going to the site another comparable night and seeing how visibility is to a human pair of eyes.
- Symbiote 9y agoThis is what the people who investigate railway accidents in the UK do. They would go to exactly the site, with an identical car, someone wearing the woman's clothing (or identical clothing) and an identical bicycle, and reenact the scene, measuring light levels and so on as the car approaches. For example, page 23-24 of this report positions a tram at various distances to check the visibility of the headlights: https://assets.publishing.service.gov.uk/media/547c8fbfed915d4c10000143/R192013_130926_Bayles_and_Wylies.pdf https://assets.publishing.service.gov.uk/media/547c8fbfed915... This is much, much more than is done for a road accident, but would seem entirely proportionate for a robot car accident investigation at this stage.