Computer Vision – Relevance and Applications
There is a widespread misconception that Computer vision and Image recognition is roughly the same thing, while in fact the truth cannot be any further. Image recognition is just one of the features computer vision encompasses. Of course the technique of image recognition is a very integral part of machine learning which in turn is the corner stone of what computer vision technology is. Even though the idea of CV dates back to the 1960s, it is only very recently that the technology can be made utilized in its full potential and this is thanks to the advancements in computation, digital information storage and higher machine learning.
Therefore you can say that it is in the toddler stage or in other words comparing the potential of computer vision to the average life span of a human, CV is only at the toddler stage. This means that in the long run the applications of CV are near limitless and even in such an early stage computer vision has successfully managed to being used across daily life and business to conduct all kinds of functions.
In a world where people take hundreds of photos just while they are having lunch has increased the digital image count online to an unfathomable amount. Studies show that it would take a single human 10 years to go through all the images that have been posted in Snapchat in the last hour and this count is slowly but surely increasing every day. To put this in perspective, this is one hour from one application that produced so much visual content. The time for human capabilities to manage such a vast database is long gone and while for a good section of time these images served no real purpose but short term entertainment, they can now be made useful for the future. We have already seen the budding applications of CV in Facebook and Google photos as one does not need to manually tag the subjects in the images. The algorithms automatically recognize the people.
The natural question that popped up in your mind would be without these digital images wouldn’t CV be redundant? One cannot think off a futuristic scenario where a computer vision relation application cannot be made useful
Automatic cars would have to understand what the obstacles in front of them are and grasp their severity to make the optimum choices. It allows any computer controlled machine, vehicle or equipment that is automatic or semi- automatic to run more efficiently and safely. 90% of all medical data is image based and there is no doubt that deriving a solution will be faster in a computer than by any medical professional in the long run. Rescue operations would be more effective when the drones or robots that are being sent can identify human personnel through the smoke, dust and mess and one does not endanger more human lives for the process of rescuing other human lives. Repair work for almost anything becomes considerably easier and can be performed by even the uninitiated and untrained as the parts that are damaged or need replacement can be identified and instructions on how to repair and edit them can also be overlaid onto the machinery itself. Another field of technology that would go hand in hand with CV would be that of Augmented Reality. The successful integration of Computer Vision in Augmented Reality is what the future hold. Similar to how smart phones are a part and parcel of our current daily life, in a decade or so AR and CV enabled devices would also be extremely common.