Installers must follow the instructions carefully for the best results.
Acting on intelligent video analytics through footage captured from traffic surveillance cameras is one of the best methods to ensure safe traffic and vehicle management. The number of security cameras is on the rise in most parts of the world as devices become more and more affordable, and authorities more aware of their benefits.
But setting up a camera that can get the best video analytic solutions requires careful consideration of certain aspects. The basic rule that applies to video analytics is that if a human eye can detect something in the footage, the video analytic system can do it too. Nevertheless, some critical factors decide the quality of footage captured.
Camera's distance from an object
Deciding where to install a camera is essential to get the best results. Any mistake in this could prove costly. Daniel Stofan, CEO of GoodVision, explains in a blog post that four crucial aspects come into play when deciding the position of a camera. Of these, distance, height, and obstacles are significant.
"To achieve the maximum guaranteed accuracy, make sure the objects to be monitored are dimensionally at least 5 percent of the scene size," Stofan said. “That meaning vehicle length should be around 60 pixels on 1280px x 720px scene. Smaller objects might be harder to detect in some cases (this is also affected by other factors — lens, blur, etc.). Also, make sure the objects are not covering a substantial portion (around 33 percent) of the scene, or it will be considered as false. How to solve this? Zoom or move."
Height and obstacles
Once the distance is sorted out, height is the next critical factor. If the camera is positioned too low, it will only detect vehicles in the front. If it’s too high, functions like number plate recognition would be challenging. Stofan suggests if your camera has standard 2.8 lenses equipped, place it between 5 and 25 meters above the ground depending on how broad space you want to monitor.
Obstacles blocking the field of vision is another vital factor to consider. This could be large objects like trees and buildings or smaller like poles and wires. If an obstacle covers a vehicle, the system may consider it as a new object when it appears again, disrupting the trajectory.
There is also the tilt factor. If you have got the distance and height right, figuring out the angle of tilt should not be difficult.
Lens and light
We will discuss the hardware requirements of video analytic systems in a separate article, but there are a few other factors related to the lens that needs to be addressed. Dirty lens or lens with scratches on them will cause blurry images and will make it difficult for the system to detect contours. Raindrops could act like physical obstacles or distort images. Image deformation can also happen because of barrel distortion, while frontal light causing flare or reflection can reduce the clarity of the picture.
"Scene lighting plays an important role in video analytics as well, however, modern systems are trained to recognize even objects in the dark," Stofan explains. "The only condition is that the objects must be at least a little bit illuminated to be visible in the image with the naked eye."
The bottom line
Jermaine Santoya, Industry Marketing Manager at Genetec,
summed up the major concerns by explaining a common experience. Often when there is a complaint of an analytic solution not functioning the way it should after installation, it would be because the installer and the customer did not follow the vendor's instructions. For instance, the camera would be installed too high or too low, or a camera that was meant to be installed indoors would be installed outdoors. So customers should understand the instructions and follow them properly for the best results.