Definition and formation of pixel noise

Resolution is used to measure how much of a bitmap image data quantity

parameters, usually expressed in pixel.In short, the resolution is the ability

of a camera to interpret images, which is the number of camera image sensor.

The maximum resolution is the capability of a camera to interpret images.,

that is the highest number of pixels of the camera.The picture we see is

composed of multiple pixels(box) as shown in figure 1-4.

                     （ Figure1）                                                                  （ Figure 2）

          
                       （ Figure 3）                                                                （ Figure 4）

When we enlarge a photograph, we can more or less find some shouldn't

exist foreign pixels, this is what we say point noise.

What is the definition of a noise?Noise is how to form ?How to eliminate the noise?

In the following, we first explore the process of camera iimaging.  

As shown in the above, camera collected the light and transfer to the sensor.

Photosensitive material on the sensor prosess the light information and then trasnfer

to the DSP(Digital signal prosess).DSP complete the photovoltaic (A/D conversion).

However, due to the photosensitive material on thesensor is too sensitive, even in a

dark environment, it also can sense some light.As you know, the light human eye

can see is limited,.Human can only see the light of wavelength between 380 and 

780.So,even in the dark environment, there are a lot of light we can't see. As long

as the photosensitive material on the sensor can perceive the light  whether or not

the human eye can see,itwill converts the light into the signal,.This result the noice

inthe image!

How is noise generated? What are the reasons?                 

1. long exposure time of noisy image   

This most appear at night, the darkness of the night sky in the image will become

the bright spot.The reason is that the charge-coupled device (CCD charge coupled

device)cannot handle a slower shutter speed brought about by the huge workload,

which led to some particular pixel out of control.In order to prevent this kind of

image noise, in the part of the digital camera is equipped with a known as the

function of "noise reduction".If you use the noise reduction function,before recording

the image,it will use digital image processing method to eliminate image noise,So

it will take a little exrea time before save the image!

2.image of JPEG format compression results the noise   

Because of the JPEG image still seem to be very natural after reducing the image

size ,you can use special methods to reduce the image data.At this point, it will

process the imgage in more than  8 * 8 pixels around as a unit.So especially

between the edge in 8 x 8 pixels and the next8 x 8 pixel unit,the combination

will not be natural.Image Noise generated by compression of JPEG formats is

also known as Mosaic Noise.The higher the compression rate, the more obvious

the image noise.Although after the reduction the image, the noise will be look

not to come out,after enlarging,printing and color compensation,noice will be

very obvious. This kind of  image noise can be solved asmuch as possible

through the use of high quality or usingother method except  JPEG format to

record images. 

3. image noisy by the fuzzy filtering

Image noise caused by Fuzzy filter is same as JPEG, which is courseed during the

processing of image.Sometimes it is coursed in the process of digital camera internal

processing, sometimes in the process of using image retouching software.For small

size image, in order to make the imagemore clear and emphasize its color edge, it

will produce image noise. Clear processing is the digital camera function of 

emphasizing on the edge of the image color and image editing software"fuzzy filter

(Unsharp Mask)"function. In a different style of digital camera also have some function

for emphasising the edge of the whole image color.After processing the image will

have other colorline on the outside ofthe edges.

So, how to eliminate the noise?

It is almost impossible.We can only improving the material quality, improve the

quality of imagingalgorithm but can not completely eliminate noise.So far, people

have developed so me of image quality evaluation method, noise is one of the

evaluation of the project.Some camera, mobile phone manufacturer also aware of

the importance of image quality, spent big price to do image quality evaluation,

from the direction of product improvement.