Digital Data Acquisition

Department of Software Engineering

Digital Data Acquisition

• The first tasks in creating multimedia content using text,

audio, video, and images is to record these individual media
types into a digital form, making it is easy to combine and
assemble these heterogeneous entities

• In the camera, there could be an image sensor CCD (charge

coupled device) array. Each sensor releases an electric
charge that is proportional to the amount of light energy falling
on it; the more the energy, the higher the charge (within a
range). The released charge is then converted into a digital
representation in terms of bits, which are ultimately used to
display the image information on a rendering device
Digital and Analog Signals

• A signal is analog if it can be represented by a continuous

function. For instance, it might encode the changing amplitude
with respect to an input dimension(s).

• Digital signals, on the other hand, are represented by a

discrete set of values defined at specific (and most often
regular) instances of the input domain, which might be time,
space, or both.

• Analog signals can have an infinite number of values in a

range; digital signals can have only a limited number of
values.
Digital and Analog Signals

Example

• A signal is analog if it can be represented by a continuous

Digital and Analog Signals

Example
Digital and Analog Signals
Advantages of Digital Signals

• In the digital medium, we can access each unit of information,

for example, it is easy to access a pixel in an image, or a
group of pixels in a region or even a section of a sound track.
(image enhancement, noise reduction)
• Stored digital signals do not degrade over time or distance
as analog signals do.
• Digital data can be efficiently compressed and transmitted
across digital networks
• It is easy to store digital data on magnetic media such as
hard drives, or solid state devices, such as flash drives,
memory cards, and so on. This is because the representation
of digital data, whether audio, image, or video, is a set of
binary values, regardless of data type.
Analog to Digital Conversion

• The conversion of signals from analog to digital occurs via two

main processes: sampling and quantization.
• The reverse process of converting digital signals to analog is
known as interpolation
Analog to Digital Conversion
 Analog to Digital Conversion
1. Sampling
• Data sampling is a statistical analysis technique used to select,
manipulate and analyze a representative subset of data points
to identify patterns and trends in the larger data set being
examined.
 Analog to Digital Conversion
1. Sampling
• We know that frequency (f) and period (t) are f=1/t

• When the signal might be under sampled, leading to artifacts,

• when the signal is over sampled, the signal requires large

amounts of storage, which might be redundant

• Practical sampling involves averaging, either in time or space.

Therefore, sampling is always associated with filtering.
Analog to Digital Conversion
1. Sampling

• Averaging
 Analog to Digital Conversion
1. Sampling
• According to the
Nyquist
theorem, the
sampling rate
must be at least
2 times the
highest
frequency
contained in the
signal
Analog to Digital Conversion
2. Quantization
• Quantization deals with encoding the signal value at every
sampled location with a predefined precision, defined by a
number of levels. In other words, now that you have
sampled a continuous signal at specific regular time
instances, how many bits do you use to represent the value of
signal at each instance? The entire range of the signal is
represented by a finite number of bits.
Analog to Digital Conversion
2. Quantization
Analog to Digital Conversion
2. Quantization
Analog to Digital Conversion
2. Quantization
• quantization effects in two dimensions for images. The results
show that the error increases as the number of quantization
bits used to represent the pixel samples decreases
Analog to Digital Conversion
2. Quantization
• quantization effects in two dimensions for images. The results
show that the error increases as the number of quantization
bits used to represent the pixel samples decreases
Analog to Digital Conversion
2. Quantization
• quantization effects in two dimensions for images. The results
show that the error increases as the number of quantization
bits used to represent the pixel samples decreases
Bit Rate

• Bit rate is of critical importance when it comes to storing a

digital signal, or transmitting it across networks, which might
have high, low, or even varying bandwidths. Bit rate, which is
measured in terms of bits per sec, consists of the following:
Bit Rate
Sampling Theorem and Aliasing
Nyquist Theorem
• In the first case (in figure) where the signal is not changing
(and, hence, has zero frequency), one single sample will
suffice for this purpose. Two samples are required for the
second signal and many more for the third one. As we go from
left to right, the frequency content in the signals increases,
and, therefore, the number of samples needed during the
digitization process also goes up.
Sampling Theorem and Aliasing
Nyquist Theorem
• The relationship between signals and sampling rate was
established by Henry Nyquist. This relationship states that the
signal has to be sampled using a sampling frequency that is
greater than twice the maximal frequency occurring in the
signal.
• For example, if the signal has a maximal frequency of 10 kHz,
it should be sampled at a frequency greater than 20 kHz.
Here, 20 kHz is known as the Nyquist sampling frequency
for that signal.
Sampling Theorem and Aliasing
Nyquist Theorem
• What happens if your sampling frequency is higher than
your Nyquist frequency?
• The answer is: nothing special.
• When it comes to reproducing your analog signal, it is
guaranteed to have all the necessary frequencies and, hence,
the same signal is reproduced. However, by performing such
over sampling, you have generated more digital data than
necessary, which increases your storage and transmission
overhead.
Sampling Theorem and Aliasing
Nyquist Theorem
• What happens if your sampling frequency is lower than
your Nyquist frequency?
• In this case, you have a problem because all the frequency
content is not well captured during the digitization process. As
a result, when the digital signal is heard/viewed or converted
back into analog form, it does not correspond to the initial
starting analog signal. This results in artifacts, which is termed
as aliasing.
Sampling Theorem and Aliasing
Aliasing
• Aliasing is the term used to describe loss of information
during digitization.
• Such undesirable effects are experienced for 1D signals such
as sound, 2D signals such as images and graphics, and even
in 3D signals such as 3D graphics.
 Sampling Theorem and Aliasing

Aliasing

• 2D input image signal. The top left shows the original signal and the remaining three
show examples of the signal reconstruction at different sampling resolutions. In all
cases, the output does not match the input because the sampling resolution falls
below the Nyquist requirement.
 Sampling Theorem and Aliasing
Aliasing
• In the one-dimensional case, the original sinusoid was sampled
(shown in gray) at a lesser frequency than the Nyquist rate.
When the samples are interpolated to reproduce the original,
you can see that the two signals do not match, except at the
exact sample points.
 Filtering

• The sampling theorem states sampling requirements to

correctly convert an analog signal to a digital signal. Knowing
the input analog frequency range can enable this conversion.
• In practice, however, the range of frequencies needed to
faithfully record an analog signal is not always known
beforehand.
• As a result, a universal sampling rate cannot be derived.

Example

• For instance, the human voice does not contain frequencies

beyond 4 kHz because of the physical or structural limitations
of the larynx.
 Filtering
Example
• However, when voice is recorded, the signal coming into the
microphone does contain higher frequencies, which are caused
by noise from the environment.
• The maximal frequency content of the signal to be recorded
might be considerably higher than 4 kHz.
• In such cases, adding a filter prior to sampling ensures that
the frequencies above a certain cutoff limit are eliminated from
the signal.
• For recording human voice, an analog filter with a 4 kHz
cutoff frequency is introduced in between the microphone and
the recording device.
• The resulting signal is then sampled at 8 kHz
 Filtering
Example
• It should also be noted that sampling alone is not physically
realizable, but is always accompanied by filtering.
• In images, light falling on a photoreceptor is averaged over the
area of the photoreceptor during a small amount of time.
Whether filtering is done prior to sampling, or as a postprocess
is not important, as these two operations are convolutions and,
therefore, commute.
• Analog filtering techniques are commonly used to capture a
variety of commonly used signals such as audio and images.
However, in the digital world, digital data manipulations also
require filters.
 Filtering
Digital Filtering
• the function of a filter is to remove unwanted parts of the signal,
such as random noise and undesired frequencies, and to
extract useful parts of signal, such as the components lying
within a certain frequency range
• An analog filter uses analog electronic circuits made up from
components such as resistors, capacitors, and operational
amplifiers (op-amps) to produce the required filtering effect.
• A digital filter uses digital numerical computations on sampled,
quantized values of the signal
Filtering
Digital Filtering
 Filtering
Digital Filtering
• Both digital and analog filters can be classified into three
categories: low-pass filters, band-pass filters, and high-
pass filters.
• Low-pass filters remove high frequency content from the input
signal. Such filters are used to avoid aliasing artifacts while
sampling.
• High-pass filters, on the other hand, remove the low-
frequency content and are used to enhance edges and sharpen
an image.
• Band-pass filters output signals containing the frequencies
belonging to a defined band. Examples are audio equipment,
WIFI routers
 Filtering
Advantages of Digital Filtering
• A digital filter is programmable
• Digital filters are easily designed, tested, and implemented on a
general-purpose computer or workstation
• Digital filters can be combined in parallel or cascaded in series
• extremely stable with respect both to time and temperature