This document summarizes a presentation on pedestrian detection using thermal imaging. It begins with an introduction to pedestrian detection and its importance for automotive safety applications. It then discusses previous work that used radar sensors, stereo cameras, and LIDAR to detect pedestrians. The document focuses on the methodology of using thermal imaging for pedestrian detection. It explains how thermal cameras work by detecting infrared radiation and temperature differences between objects. It also defines the different infrared spectrum bands and discusses how thermal imaging has advantages over visible light cameras for human detection applications. In conclusion, it states that thermal imaging outperforms visible light but more research is needed to fairly compare the two methods.

1. WELCOME

2. Presented By:
Hassan Iqbal
Fahad Saeedi
Student of BSCS
Computer Science Department NCBA&E

3. Pedestrian Detection

4. Introduction
• Pedestrian detection is an essential and significant task in
any intelligent surveillance system, as it provides
fundamental information for the semantic understanding of
any scene.
• It has an obvious extension to automotive applications
because of the potential to improve safety systems.

5. Motivation
• Pedestrian safety methods are improving through different
ways and these can be summarized as follows : cautionary
signals, emergency alarm, auto braking and deploy collision
mitigation.
• Most accidents occur because of the visibility problem, such
cases can be improved by using radar waves which can be
attached to the bumper side on the vehicles.

6. Importance
• This project is mainly motivated based on the increased need
to protect pedestrians from road accidents
• Pedestrian visibility can be increased by improving the road
lighting since most pedestrian injuries occur at night.

7. Literature
• Steffen Heuel and Hermann Rohling developed a
classification algorithm for automotive application using
radar sensors (at 24GHz), which can be used for measuring
velocity and distance with a band-width of 150MHz [8]. The
paper proposed two systems such as single radar system,
that measures the transmitted signal using a single MFSK
(Multi-Frequency Shift Keying) at 39ms

8. Continue
• Gavrila and Munder, [9] proposed PROTECTOR system (a
real-time stereo system for pedestrian detection and
tracking). The highlights of the method is that it used a
texture based classification and the method used fixed
cameras at 25 meters apart. The pictures are stored in
frames, which resulted in 71 percent pedestrian detection
and 0.1 false alarms/frame

9. Continue
• Cristiano Premebida et.al. [11] used 3D laser sensors
commonly known as LIDAR in the detection and evaluation
for depth perception of road crossing pedestrians [17]. There
has been 3741 frames used for the detection on which 52%
were detected successfully. The con is that the pedestrian
detection algorithm resulted with less accurate values.

10. Methodology
THERMAL IMAGING
• Thermal imaging can be seen as a method of improving
visibility of objects in a dark environment by detecting the
objects' infrared radiation and creating an image based on
that information. Here's an explanation of how thermal
imaging works:
• All objects emit infrared energy (heat) as a function of their
temperature.
• The infrared energy emitted by an object is known as its
heat signature.

11. Continue
• In general, the hotter an object is, the more radiation it
emits.
• A thermal imager (also known as a thermal camera) is
essentially a heat sensor that is capable of detecting tiny
differences in temperature.
• The device collects the infrared radiation from objects in the
scene and creates an electronic image based on information
about the temperature differences.

12. Continue
• Because objects are rarely precisely the same temperature as
other objects around them, a thermal camera can detect
them and they will appear as distinct in a thermal image.
• Thermal images are normally grayscale in nature: black
objects are cold, white objects are hot and the depth of gray
indicates variations between the two images.
• Some thermal cameras, however, add color to images to help
users identify objects at different temperatures.

13. Continue
• Nowadays this technology has contributed in many areas
and in this paper an investigation about its contribution in
the field of pedestrians’ detection and crowd counting.

14. Sample of gray level thermal image

15. INFRARED BANDS AND THERMAL SPECTRUM
• In Latin ‘infra’ means "below" and hence the name 'Infrared'
means below red. ‘Red’ is the color of the longest
wavelengths of visible light.
• Infrared light has a longer wavelength (and so a lower
frequency) than that of red light visible to humans, hence
the literal meaning of below red.
• 'Infrared' (IR) light is electromagnetic radiation with a
wavelength between 0.7 and 300 micrometers, which
equates to a frequency range between approximately 1 and
430 THz.

16. Continue
• IR wavelengths are longer than that of visible light, but
shorter than that of terahertz radiation microwaves.
• Objects generally emit infrared radiation across a spectrum
of wavelengths, but only a specific region of the spectrum is
of interest because sensors are usually designed only to
collect radiation within a specific bandwidth.

17. Continue
• As a result, the infrared band is often subdivided into smaller
sections.
• The International Commission on Illumination (CIE)
recommended the division of infrared radiation into three
bands namely, IR-A that ranges from 700 nm–1400 nm (0.7
µm – 1.4µm), IR-B that ranges from 1400 nm–3000 nm (1.4
µm – 3 µm) and IR-C that ranges from 3000 nm–1 mm (3 µm
– 1000 µm).

18. Continue
• A commonly used sub-division scheme can be given as
follows: Near-infrared (NIR, IR-A DIN): This is of 0.7-1.0 µm in
wavelength, defined by the water absorption, and commonly
used in fiber optic telecommunication because of low
attenuation losses in the SiO2 glass (silica) medium.
• Image intensifiers are sensitive to this area of the spectrum.
Examples include night vision devices such as night vision
camera.

19. Continue
• This is of 13 µm. Water absorption increases significantly at
1,450 nm. The 1,530 to 1,560 nm range is the dominant
spectral region for long-distance telecommunications.
• Mid-wavelength infrared (MWIR, IR-C DIN) or Intermediate
Infrared (IIR): It is of 3-5 µm. In guided missile technology the
3-5 µm portion of this band is the atmospheric window in
which the homing heads of passive IR 'heat seeking' missiles
are designed to work, homing on to the IR signature of the
target aircraft, typically the jet engine exhaust plume.

20. Continue
• Long-wavelength infrared .
• This infrared radiation band is of 8–14 µm.
• This is the "thermal imaging" region in which sensors can
obtain a completely passive picture of the outside world
based on thermal emissions only and require no external
light or thermal source such as the sun, moon or infrared
illuminator.
• Forward-looking infrared (FLIR) systems use this area of the
spectrum.

21. Continue
• Sometimes it is also called "far infrared“. Very Long-wave
infrared (VLWIR): This is of 14 - 1,000 µm.
• NIR and SWIR is sometimes called "reflected infrared" while
MWIR and LWIR is sometimes referred to as "thermal
infrared".
• Now, we can summarize the wavelength ranges of different
infrared spectrums as in Table.

22. Table
Wavelength range for different spectrums.

23. Conclusion
• From the previous discussion it is clear that dealing with
thermal bands don’t need any special techniques for
processing:
• 1) Edge detectors: (Ex: Sobel filters).
• 2) Morphological operators.
• 3) Training classifiers: (Ex: Ada-boost & Bayesian).
• 5) Finding interest points and region of interests.
• 6) Features matching

24. Continue
Recent researches proved that thermal imaging has
outperformed visible bands in the field of human detection
plus that it allowed the presence of many applications that are
needed in many different fields nowadays. However, there
still a lake for researches that introduce a fair comparison
between the two bands that may introduce challenges of this
new approach.

25. .
Thank You