1. Imaging Science Fundamentals
Digital Imaging: CCDs
Imaging Science Fundamentals

2. Charge Coupled Device (CCD)
Light Sensitive Area
CCD replaces AgX film
Based on silicon chip
Disadvantages vs. AgX:
Difficulty/cost of CCD manufacture; large arrays are VERY expensive
“Young” technology; rapidly changing

3. Response of photographic
Response of CCD
The response of CCD is linear (i.e., if 10,000 captured photons corresponds to a digital count of 4, then 20,000 photons captured yields a digital count of 8)
Linearity is critical for scientific uses of CCD
Log H
Density
Response of photographic
negative
Exposure
Digital Count
Response of CCD

4. Spectral Response (sensitivity) of a typical CCDUV
Visible Light IR
Relative
Response
300
400
500
600
700
800
900
1000
Incident Wavelength [nm]
Response is large in visible region, falls off for ultraviolet (UV) and infrared (IR)

5. Goal of CCD
CCD
Photons
Electronic Signal
Capture electrons formed by interaction of photons with the silicon
Measure the electrons from each picture element as a voltage

6. Spatial Sampling
Scene
Grid over scene
Spatially sampled scene
Since the CCD is divided into a small segment called pixels, the captured image is made up of these little pixels. Sometimes if a digitally captured picture is enlarged, the small details look “blocky”; This is due to the fact that the continuous scene was spatially sampled by the CCD. If the number of pixels is increased (and the capture area remained constant) the blockiness become more and more unresolvable to human perception.
This doesn’t mean that one should sample the scene with the smallest grid all the time- The image becomes larger and larger, requiring more for storage. So a scene should be captured so that it has enough resolution for its purpose; anything else would be waste of space and time.
(You wouldn’t use a sledge hammer to drive a tack into the wall, would you?)
When a continuous scene is imaged on the array (grid) formed by a CCD , the continuous image is divided into discrete elements.
The picture elements (pixels) thus captured represent a spatially sampled version of the image.

7. Basic structure of CCD Image Capture Area Rows Voltageout Columns
Divided into small elements called pixels (picture elements).
Shift Register
Image
Capture
Area
Rows
CCDs are made basically of three main parts.
Image capture areas are where the photons are captured and stored as electrons. It is divided into small discrete area called pixels, which stands for picture elements. Each pixels are identical in structure, and the number of pixels determine the spatial resolution or how fine the picture will look.
Depending on the application in which the CCD is being used, a variety of sizes can be found. A camcorder and digital cameras often use 1/4inch to 1/3 inch CCDs with 640 rows by 480 columns, while a scientific grade CCDs can have 4000 rows by 4000 columns!
A Shift Register or MUX functions as a reservoir tank to hold the electrons during readout. This is also called horizontal shift registers.
the voltage accumulated within each pixel is very small, so in order to be useful, it must be amplified by a preamplifier to get it in a voltage where it can be quantized by a digitizer.
Now let’s take a look at the structure of the image capture area in more detail.
Voltageout
Columns
preamplifier

8. Magnified View of a CCD Array
Individual pixel element
CCD
Here you can see the stucture of a real CCD. Notice that the gates are overlapping one another. This is done purposefully to improve the performance of CCD.
Close-up of a CCD Imaging Array

9. CCDs as Semiconductors
Insulator
Conductor
Conductors allow electricity to pass through. (Metals like copper and gold are conductors.)
Insulators do not allow electricity to pass through. (Plastic, wood, and paper are insulators.)
Some materials are halfway in between, and are called semiconductors.

10. Basic structure of a pixel in a CCD
Metal gate
Oxide
Layer
Silicon base
Here we have a cross section of a CCD image capture area.
In review, something that lets electricity pass through is called a conductor.
Something that blocks electricity from flowing is called a insulator.
CCD is made on a silicon wafer using microlithographic techniques.
Silicon is an element with a unique property; It is neither insulator or a conductor. In certain conditions, it acts as a conductor, while in other conditions, it acts as a insulator. This is called a semiconductor.
On top of the silicon is a oxide layer which functions as an insulator. This layer is made essentially by “rusting” the surface of the silicon inside a temperature controlled oven.
And on top of that is metal gates, which is a conductor, freely passing electricity.
One pixel
Silicon is a semiconductor.
Oxide layer is an insulator.
Metal gates are conductors.
Made with microlithographic process.
One pixel may be made up of two or more metal gates.

11. Photon/Silicon Interaction
The most important reason that silicon is used in CCD is because of how photon interacts with silicon crystal.
When a photon strikes the silicon, it penetrates into the crystalline matrix. And if the conditions are right, it knocks off an electron from one of the silicon atom.
The electron wanders around the silicon matrix, in a random manner.
Eventually the electron gets absorbed by the silicon, returning to the ground state or the normal state.
Photon knocks off one of the electrons from the silicon matrix.
Electron “wanders around” randomly through the matrix.
Electron gets absorbed into the silicon matrix after some period.

12. Collection stage Voltage
Recall the structure of the CCD, where it is made of the silicon base, oxide layer as an insulator, and the metal gates. When a voltage is applied to these metal gates, it moves the electrons around in such ways that an area where it has no electrons form. This region is called a depletion region.
This depletion region is the light sensitive area, where the electrons formed from the silicon/photon interaction gets captured.
Voltage applied to the metal gates produces a depletion region in the silicon. (depleted of electrons)
Depletion region is the “light sensitive” area where electrons formed from the photon interacting with the silicon base are collected.

13. Collection stage Voltage e- e- e-
Now let’s take a look how an image is captured by the CCD.
The light from the object is imaged onto the CCD. The photon penetrates the metal gate and the oxide layer and is absorbed by the silicon base. An electron is formed from the absorption, where it wanders around the matrix randomly. If it happens to wander into the depletion region, then the electron is captured and stored in the region. Without this region, the electron recombines with the rest of the matrix after some time.
Note that the metal gate next to the active collection area is not on, even though it is a part of the pixel. This gate is used for transporting the charge collected. The readout involves a systematic movement of the charge from one pixel to the next.
Electron formed in the silicon matrix by a photon.
Electron wanders around the matrix.
If the electron wanders into the depletion region, the electron is captured, never recombining with the silicon matrix.

14. Collection
Light e- e- e- e- e- e- e- e- e- e- e-
The number of electrons accumulated is proportional to the amount of light that hit the pixel.
There is a maximum number of electron that these “wells” can hold.

15. Readout
Now that the electrons are collected in the individual pixels, how do we get the information out?
Alright! How do we
get the electrons
out?!
At this stage, we have electrons representing how much light hit that particular pixel, and are stored in the depletion region of each pixel. How do we read them out so that we can get the information captured efficiently?

16. Readout
How do you access so much data efficiently? (i.e. a 1024 x 1024 CCD has 1,048,576 pixels!)
Possible solutions:
1. Have output for individual pixels.
Too many “wires”
2. Somehow move the charges across the CCD array and read out one by one.
Bucket Brigade
So how would you read out that much information is a short amount of time? CCD are used in video cameras as well as digital still cameras, where the image must be read out in 1/30 of a second, as we’ll learn about TV in later section.
There are several possible solutions. The obvious one is to connect each and every one of the pixels in the CCD and read them out individually and simultaneously. This is fast; however, you need a lot of “wires” which is hard to do in microlithography. Although this kind of approach is used in a similar device called CID or Charge Injection Device.
Second possibility is to actually move the accumulated charge across the CCD array using a systematic transfer of charge from one pixel to the next. This method is called bucket brigade and used most of the time in one form or another.

17. Bucket Brigade
By alternating the voltage applied to the metal gates, collected electrons may be moved across the columns.
Bucket brigading across the pixels works with alternating the voltage applied to each of the gates. Recall that the voltage applied produces a region in the silicon where the electrons are attracted to. When the gate voltage is slowly adjusted, the charge migrates from one region to to the other like passing a bucket. If we were to faithfully follow the bucket brigade analogy, the physical “buckets” are never passed from one pixel to the other; rather, the water inside the bucket is passed on from one to the other by dumping the content to the next bucket.
This is why one pixel has more than one metal gate; In fact, some CCD has up to 4 metal gates per pixel! These are referred to as “Four phase” CCD. In our case, since we have only two gates per pixel, it is a “Two phase” CCD. e- e- e- e- e- e- e- e- e- e- e- e- e- e- e- e-

18. Bucket Brigade
Charge is marched across the columns into the shift register, then read out 1 pixel at a time.
100 transfers
200 transfers
Shift Register
100 pixels
When the bucket brigade reaches the end of the row, it must be then transferred down the column. For this, a special portion of the CCD called the shift register is used. Shift register does not collect any electrons during the capture stage. It is only used for moving the electrons down to the preamplifier at the end of the bucket brigade.
An entire column is transferred to the shift register, then the shift register bucket brigades down the column. Shift register must work 100 times as fast (in this case) since there are 100 pixels in the vertical direction. If this CCD had 1000 pixels in each direction, then the register must work 1000 times as fast.
Though this bucket brigade seem to work well, it has some limitations that degrade the signal with each pass. So the pixel farthest away from the readout preamp gets degraded. Kinda like how with each dumping of the water from one bucket to the other, some droplets are left behind. When the the total drops lost are measured, then it could end up with a significant amount.
100 transfers
1 transfer
100 pixels

19. Converting Analog Voltages to Digital
Analog voltage is converted to a digital count using an Analog-to-Digital Converter (ADC)
Also called a digitizer
The input voltage is quantized:
Assigned to one of a set of discrete steps
Steps are labeled by integers
Number of steps determined by the number of available bits
Decimal Integer is converted to a binary number for computation
ADC
6.18 volts
(117)

20. Bits and Bytes
In the digital domain, there are only two possible numbers in a digit: 0 or 1.
This numbering system is called a binary system.
Each digit is called a bit (Binary digIT).
Byte is 8 bits
Decimal 1 2 3 4 5
Binary 1 10 11
100
101
Bit is a part of a counting system called binary system where the only possible values are 0 or 1. This is convenient for computers and other digital stuff since computers are essentially a whole bunch of small switches. And these switches can only be on (1), or off (0).
Each of the digits are called a bit (from Binary digIT). Just as a decimal system when the digit carries over one place when the number increases from 9 to 10, in a binary system, the digit carries over when it increases from 01 to 10, or 2 in decimal system. So number 255 is in binary.
Also, 8 bits is also called a byte.
Subsequently, 4 bits is called a nibble, but this is seldom used.

21. Bits Bits dictate how fine the quantization levels are.
An n bit system can represent 2n numbers.
1 bit system = 21 = 2 levels (“Black” or “White”)
Binary number has a maximum possible value just as any number system.
(How many numbers can you count to with your fingers? If you use the decimal system, you can only count to ten, but if you use the binary system, you can count up to 1,023!)
As a rule of thumb, an n bit system can represent 2^n numbers.
This number of bits dictate how fine the quantization levels are in the ADC
8 bit system = 28 = 256 levels
12 bit system = 212 = 4096 levels

22. Quantization ADC 6.8 volts 174.08 6.8v
Let’s say our 8 bit ADC accepts input voltage range of 0 to 10v.
Volts DC
10v
172
173
174
175
176
255
Since there are 256 discrete levels in an 8 bit system, each level will be 10v/256 or volts per analog-to-digital unit (ADU).
6.8 volts
174.08
6.8v
So, if the input voltage was 6.8 volts . . .
This is how a ADC would quantize an input and come out with output.
Let’s say that our system is equipped with an 8 bit ADC which accepts input voltage range of 0 to 10 volts 0v
Since ADU are stored as binary integers, the decimal must be truncated (to 174).
6.8 volts/ volts per DC =
Binary equivalent of 174 is

23. Quantization 25 40 64 97
150
Spatially sampled scene
Numerical representation
Now with this ADC, we can convert the collected electrons which is proportional to the number of photons which hit each of the pixel (which from the light that made up the image) to some number representing the image!
Spatially sampled image can now be turned into numbers according to the brightness of each pixel.