ENGLISH
INDONESIA
Noise if translated into Indonesian means noisy / noisy. Anything that is loud/noisy is usually not desirable. For example noise in voice or signal. A sound that should be clear becomes rustle due to noise. Well, the definition of noise on CCTV cameras is more or less the same. The difference is only the form and the cause. Noise on the camera in the form of dots that spread. There are 3 types of noise in the camera
3 Kinds of Noise on the Camera
1. Noise Salt and Papper
Dots, like sprinkling of salt, will whiten the noise-affected spots
2. Gaussian Noise
The number of colored dots in the image is equal to the percentage of noise
3. Film Grain Noise
Dots that give black color at points that are exposed to noise (like the appearance of a film in a photo)
The cause of noise in CCTV cameras is due to the lack of light falling on the camera sensor. So that the image becomes grainy and looks blurry. The Digital Noise Reduction (DNR) feature on CCTV cameras will eliminate this noise. DNR works by analyzing 2 consecutive images/frames. The DNR program routine will remove noise from the last frame/image that was not present in the previous frame/image.
Traditional DNR only removes noise on moving objects, or also known as foreground objects. Meanwhile, background noise, or noise on immovable objects such as roads or houses, is still there. Disadvantages of Traditional DNR can be overcome by 3D-DNR(3DNR). 3D-DNR works by analyzing each pixel with the other pixels that surround it. This technique is called Spatial Noise Reduction. Traditional DNR only removes noise on moving objects, or also known as foreground objects. Meanwhile, background noise, or noise on immovable objects such as roads or houses, is still there. Disadvantages of Traditional DNR can be overcome by 3D-DNR(3DNR). 3D-DNR works by analyzing each pixel with the other pixels that surround it. This technique is called Spatial Noise Reduction. With 3D-DNR, the resulting image will be cleaner because 3D-DNR removes noise not only on foreground objects but also on background objects.