We do a lot of coverage here at ReelSEO talking about video being a way to tell a story, transmit a message and expand brand audience or awareness or sales. But one of the bigger, newer uses for video is in the analytics department. Video is ubiquitous, from the surveillance cameras in retail outlets to the traffic cams on city streets and highways to smartphones, tablets, and police cruisers. A lot of that sits in some repository somewhere while much of it gets uploaded to the web. But what if it were all tapped in a somewhat sci-fi way to become an all new source of big data and analytics?
Some researchers are working on just that sort of thing. For example, take a surveillance video of a retail shop and apply a heat map algorithm to the people as they walk through and you find out all the very busy parts of the store and all the dead space. Then you can make a better path for them where they end up seeing more product and perhaps increase sales.
Do something similar to highway traffic cams and you can see which lanes cars drive in, what speed they drive at and where most accidents are caused in order to relieve traffic congestion, create better routes and prevent more collisions.
Apply it to city street cams and you can begin looking at how to mitigate commuter clutter and traffic light wait times by finding out where cars wait the longest to turn or where more cars transfer from one thoroughfare to another.
Construction and Deconstruction
Compile online video feeds during a major event and you could generate a three-dimensional version of that event. Multiple people viewing the same thing from multiple angles could be combined in order to aid disaster relief, direct first responders to those most in need and analyze crowd mentality in order to enhance their experience at festivals, concerts and other events.
Taking a video apart and using object recognition is a pipe dream for retail, or is it? For years now we have been able to tag items in videos and attach them to certain products for advertising and marketing. But what if we had an algorithm that stored images of products and items and could do it all on its very own without help from a human? Every video could potentially become an advertisement or marketing piece. Software could analyze the contents of videos as well as the context and then present offers to the creators to be paid for inclusion of particular brands or items, or the software could be used to ensure that videos are including those very items.
Maximization of Realization
Trying to determine the efficacy of placing a billboard advertisement or perhaps building a new franchise location? Get exact numbers of people or cars passing by, which direction, time of day and other information before you decide and you could find the optimal placement for anything.
Granted, there are going to be privacy concerns and issues. Companies will cross lines and consumers will file class action lawsuits until it all gets boiled to down, when and where can you expect some form of physical privacy? While in a retail outlet or a public space? You probably have much less privacy than you think already. The question is where will the line lie that becomes 'too far?' Will it be when that video footage is used to start real-time advertising to customers in a store based on their browsing or purchasing history? How is that any different than what is done online now? Amazon knows what you look at, knows what you buy, they've even put forth a patent recently that states they would probably be able to predict what you will buy and send it to you before you even buy it. They do that just from your browsing their site and from your purchasing habits. Imagine what they could do if they had brick-and-mortar shops which could watch your every move?
Potentially Dystopian Endeavors
Sounds creepy and Orwellian right? I haven't even touched on what governments could do and probably are already doing. We are a population that is watched nearly every moment of every day once we step outside our homes. I can walk out of my house and before I walk the full block I know I've turned up on the footage of at least a half dozen security cameras. Walk through any downtown area these days and surely you'll see the public cameras sitting in their little plastic bubbles being protected from the elements so that they can keep a bird's eye view of everything, not just intersections for traffic accidents and infractions.
Even in our homes video is captured and uploaded to the web. If you share that video with anyone or any entity like say, YouTube, or an online brand for a contest, etc. You can bet it will probably be picked apart and analyzed by someone sooner or later. Oh you drink Coke while you play Call of Duty? Here's a coupon. I know because I saw that video you posted on Twitch and there were several empty Coke cans next to you, or you mentioned going to grab a Coke while playing.
More Data Than All Other Media Combined
Video gives us more information per second than any other form of media or data tracking. We get not only the location, but also the situation, the participants, the brand of clothing, drink, food, even type of device used to shoot the video. We know how long something lasted, how much time you spend doing something like washing your car or walking to work.
All of this data, this video, is being accumulated and analyzed and eventually that analysis will begin to be implemented. Hopefully, the only thing it gets implemented into is advertising and marketing and public safety mitigation, not pre-crime prevention or health insurance claims or some other dystopian future, Just benevolent uses that don't make us all feel like a nation that is spied upon anymore than we already do thanks to the NSA and recent whisteblower revelations.
Good vs. Evil - We Must Choose Wisely
I don't want to leave this on a negative note because some of these new analytical uses of computer vision and digital video are good. They'll allow retailers to make better decisions in terms of what kind of stock they keep on hand, especially when thinking about perishable food items lowering overall food waste and loss. It will help retailers maximize their space making for better sales and better retail experiences for consumers. It will make for safer roads and highways, faster commutes and even better searching online for shopping and entertainment. Better walking experiences for pedestrians, safer travel for city-based bicyclists, perhaps even better farming through being able to have a computer choose what crop is perfect to pick that day and ship to market for maximum shelf life or best taste. The only thing that limits the good that can be done with all this is not doing any of it or trying to prevent it because it may be abused. Everything in the world may be abused, we just have to make sure that we all use these newfound superpowers for good and not evil.