Images are amazing. They provide an instantaneous snapshot of the world we live in from a particular point of view. In a world where cameras are so ubiquitous that just about everyone in the world has one in their pocket, it is worth taking a second to think about this deeply profound invention. With an image, you actually freeze time. You turn an instant into eternity. You create a copy of singular moment that you can revisit and share. You capture the moment, you own the moment. It is yours, and available at the press of a button.
Ok, all hyperbole aside, images are pretty cool, and yes, they really do capture an instant in time. But the real question is why do we care? In an age where everyone seems to have a camera in their hand (often with a phone embedded in it), and few are afraid to use them, understanding how images are used is fundamental to understanding why we take them. In thinking about this question, I've been reviewing my own library of photos and I've identified the following categories for why the pictures were taken. While not a comprehensive list, this framework provides context for thinking about customer needs and expectations for the images they take as a critical step towards developing imaging systems that are compelling and valuable to the user.
The photos used to illustrate these categories outlined in this post are from my personal library and provide real examples of how I used images in these different contexts.
Some of my favorite images are not conventionally good images. They may be a bit blurry, or faded with uneven exposure, but they were taken at a time and place of personal significance. An excellent example would be this photo of me feeding an elephant that seems like it was taken a lifetime ago. Objectively, it's a low quality image, but its an instant reminder of my college volleyball team, my first trip overseas, and that one time I was chased by a pachyderm. It doesn't take a pristine image to trigger these memories. As long as you see enough to remember the moment, they are good enough. These are the images we tend to treasure the most.
With the advent of the Internet and social media, sharing an experience seems to be the primary way people now use images. This is particularly the case now that everyone has a recording device in their hand. Go to any concert, sporting event, or high school musical and you will see a large percentage of the audience with their phones in the air making a video. While this may be a bit distracting for the other attendees, it allows those unable to be there in person to get a feel for what they missed.
The experiences we share are typically live or video recordings of dynamic events. In these cases, the video and audio needs to be good enough to make the viewer feel like they were present at the time of the recording, so the video quality needs to be good enough not be distracting.
Immersive experiences are on the horizon and these will transforms how experiences are shared. While the image quality for user will be driven by the human visual system, much more data will be required on the acquisition side to make the experience increasingly interactive.
When one of my kids wasn't able to attend a concert with the rest of the family, we recorded some of the show so he could see what he missed.
Photogrammetry is refers to accurately recording a scene to provides a reliable information about the physical world. With applications ranging from crime scene analysis to 3D modeling, these images provide a reliable evidence of time and place. Examples include using multiple photographs of photogrammetric techniques to generate virtual twins of real objects, forensic photography of crime scenes that can be used to establish guilt or innocence, and perhaps more common, taking photographs of complex scenes to record the relative position and orientation of objects. This is particularly useful when rebuilding complex components or when trying to compare pieces.
The quality of images required for forensics is heavily application dependent. The resolution required to know the relative location of the piers and pipes as shown on the image to the left
One of a series of images of the crawlspace of my house. I needed to know the location of the piers relative to the electrical and plumbing infrastructure.
It is not uncommon to take images for documentation purposes. This can include photos of original documents, the steps for disassembling a complex component, or recording where one parked. Documentation photos are generally taken with a prior understanding of what relevant information is at the time the picture is taken.
Images used for documentation need to have enough resolution to see the documentation in necessary detail to recreate the content, but these requirements often considerably less that forensic applications.
A picture of a map of Spisske Podhradie in Slovakia that we used while navigating the city on a recent vacation.
When looking at objects that are beyond the human visual capacity, images allow us to see what cannot be directly observed. Whether the observer is looking at microscopic objects such as bacteria, viewing distant galaxies through a telescope, or just trying to take in a large dynamic scene, images allow us to see things that we would otherwise be unable to perceive. My earlier post on Camera Array Systems: More than meets the eye is a great example of how images allow us to understand the world better than through direct observation.
For me, perhaps the most common example is when I use my phone to take a photo of fine print and the blow it up to where I can read it. The imaging system must be matched to what you are looking for whether it's cellular resolution for imaging cells or the wide field and high resolution of a security system.
The fine print on this USB charger was way too small for me to see. This photo allowed me to zoom in to figure out why my phone was charging so slow.
Understanding the different contexts in how people use images is important for developing cameras and imaging systems that are compelling and meaningful for the user. The categories and examples that I have discussed in this post are the ones that are I find useful, but this isn't intended to be a definitely list. Instead, this post is an invitation for camera and software designers to think a little more deeply about how they use images and use that as motivation for future innovation.