Introduction to

Artificial Intelligence & Robotics

03. Python Library
[Link]

Ukcheol Shin
Robot Physical AI Lab., KENTECH

Robust Physical AI School of Energy

& Robotics Lab. Engineering
Lecture’s agenda
[Python Library]
• [Recap] Important thing is a logical thinking!
• Python Library
• Matplotlib
• Word Cloud
• Numpy
• OpenCV

Robust Physical AI School of Energy

3
& Robotics Lab. Engineering
 What is Programming?

New idea! Design algorithm Write a source code Test program

Programming
“the process of designing, writing, testing, and maintaining
a sequence of instructions (a program) to solve a problem.”

Robust Physical AI School of Energy

4
& Robotics Lab. Engineering
 Important thing is logical thinking!

Write a source code

New idea! Design algorithm

Different syntax
Robust Physical AI School of Energy
5
& Robotics Lab. Engineering
Important thing is logical thinking!
# basic syntax of window CMD
# 1. standard I/O (input, print)
echo Hello, window CMD!
>> Hello, window CMD!

set /p name=Enter your name:

echo %var%

# 2. set variable
set var=DLF
echo %var%
>> DLF

# 3. if-else
set age=19
if %age% geq 18(echo Adult) else (echo cute!)
>> cute!
Window command prompt (CMD) # 4. for loop
for %i in (1 2 3 4 5) do (echo Number: %i)
>> 1 2 3 4 5

Robust Physical AI School of Energy

6
& Robotics Lab. Engineering
 What is Programming?

New idea! Design algorithm Write a source code Test program

Programming
“the process of designing, writing, testing, and maintaining
a sequence of instructions (a program) to solve a problem.”

Robust Physical AI School of Energy

8
& Robotics Lab. Engineering
 Programming
• Algorithm
- A set of rules or steps used to solve a problem
# define functions
• Function def read_list():
…
- A block of reusable code, a way of grouping related steps
def sort_list():
…

• Data Structure def print_list():

…
- A particular way of organizing data in a computer
# main body
read_list()
sort_list()
[Link] print_list()
[Link]

Robust Physical AI School of Energy

9
& Robotics Lab. Engineering
 Standard data types in Python
• A data type defines the kind or category of data (integer, float, string, etc)
• We can assign various types of data to a variable.
Python – Data Types

Numeric Boolean None Dictionary Set Sequence Type

Integer Float Complex number Strings List Tuple

Data types Examples Data types Examples Data types Examples

Integer …, -2, -1, 0, 1, ... Bool True, False Set {1, 0.3, “a”, “ab”}
Float 3.2, -0.14, 35.75 String “Hi, data!”, “123” List [1, 0.3, “a”, “ab”]
Complex 1+3j Dictionary {1:”a”, 2: “b”} Tuple (1, 0.3, “a”, “ab”)
Robust Physical AI School of Energy
10
& Robotics Lab. Engineering
 Procedural Programming
• What is Procedural Programming?
➔ Think of a program as a set of “Procedure”

• To built a program, # define functions

Q. What kinds of “Procedures” are necessary? def read_list():
…

def sort_list():
• What is procedure? …

• A block of code that performs a specific task def print_list():

• A unit of work or step in the overall program …

• Different languages use different terms # main body

: e.g., procedure (pascal), function (C, python), read_list()
sort_list()
subroutine (fortran), method (Python), etc print_list()

Robust Physical AI School of Energy

11
& Robotics Lab. Engineering
 Object-oriented Programming
• What is Object-oriented Programming?
➔ Think of a program as a set of “Object” Object Object

Data Data
• To built a program,
Q. What kinds of “Object” are necessary?
Function Function

• What is object?
Function Function
• A unit that has state (i.e., data), behavior
(i.e., function), and identity.
• Resembles real-life entities with state and
action.
• An instance of a class.
Robust Physical AI School of Energy
12
& Robotics Lab. Engineering
Organizing Program Logic in Python
As programs grow larger, we need bigger “containers” to organize logic:

• Functions: A block of code that performs a specific task.

• Objects: An entity that combines data (attributes) and behavior (methods)

• Modules: A single Python file (.py) containing functions, classes, or variables.

• Libraries: A collection of modules organized together, often distributed for reuse.

Robust Physical AI School of Energy

13
& Robotics Lab. Engineering
A lots of python libraries …

Over 100,000+ libraries!

✓ Cannot cover each library & its usage & functions & …
Not my job ^-^; that is your job!
Robust Physical AI School of Energy
14
& Robotics Lab. Engineering
 A lots of python libraries …
Data Data analyze & visualization
• Number: 10293, 37.2℃, 175cm, 70kg, …
• Text: “It is raining today”, SNS’s post, News, …
• Signals: images, videos, audio, electric signal, …

Weather
Stock price

Trajectory
Robust Physical AI School of Energy
& Robotics Lab. Engineering
A lots of python libraries …
• Pandas → Tabular data analysis (filtering, grouping, …)

• NumPy → Scientific computing, numerical operations for multi-

dimensional arrays

• Matplotlib / Seaborn → Data visualization (histogram, heat map, pie

chart, graph, …)

• Requests → HTTP communication library, used for REST API calls,

downloading files, interacting with websites

• BeautifulSoup → Web Scarping library, parses HTML/XML

documents, extract titles, links, tables, or any structured web content,
often used together with Requests
Robust Physical AI School of Energy
& Robotics Lab. Engineering
A lots of python libraries …

• Pillow → Image processing library, easy to use with Tkinter and PIL
(Python Imaging Library)

• OpenCV → computer vision toolkit for image and video processing

• Scikit-learn → Machine learning toolkit (classification, regression,

clustering, K-means)

• PyTorch/TensorFlow/ Keras → High-performance deep learning

framework

Robust Physical AI School of Energy

& Robotics Lab. Engineering
Python package installation
Pip (Python Package Installer)
• Python provides pip, the official package installer, for downloading and managing
external libraries
• Downloads packages from PyPI (Python Package Index, [Link] )
• Supports library installation, upgrade, removal, dependency handling
• Included by default in most Python distribution

In your CMD or powershell or VScode terminal

Robust Physical AI School of Energy
& Robotics Lab. Engineering
 Python package installation
Pip (Python Package Installer)
• Python provides pip, the official package installer, for downloading and managing
external libraries
• Downloads packages from PyPI (Python Package Index, [Link] )
• Supports library installation, upgrade, removal, dependency handling
• Included by default in most Python distribution
# basic installation command # check installed packages
pip install matplotlib pip list
pip show matplotlib
# installation with specific version
pip install pandas==1.5.0 # uninstall a package
pip uninstall matplotlib
# upgrade an existing package
pip install --upgrade matplotlib

Robust Physical AI School of Energy

& Robotics Lab. Engineering
Python package installation
Wait! Where is my python package located in?
• $ python –c “import sys, pprint; [Link]([Link])”

Robust Physical AI School of Energy

& Robotics Lab. Engineering
Matplotlib and word cloud

Matplotlib Word cloud

: data visualization (histogram, : generates visual “word clouds”
heat map, pie chart, graph, …) from text frequency
Package: [Link] Package: [Link]
Source: [Link] Source: [Link]
Document: Document:
[Link] [Link]

Robust Physical AI School of Energy

& Robotics Lab. Engineering
Matplotlib
$ pip install matplotlib wordcloud
> On your window CMD, powershell, VScode terminal

Robust Physical AI School of Energy

& Robotics Lab. Engineering
Matplotlib
Matplotlib
: one of the most widely used Python plotting libraries
> Line plot, bar charts, Pie charts, 3D plot
> 1D data, 2D image, 3D data, …

Robust Physical AI School of Energy

& Robotics Lab. Engineering
 Matplotlib
Matplotlib
: one of the most widely used Python plotting libraries
> Line plot, box plot, bar charts, Pie charts, 3D plot
> 1D data, 2D image, 3D data, …
import [Link] as plt Line plot
X = [ "Mon", "Tue", "Wed", "Thur", "Fri", "Sat", "Sun" ]
Y = [15.6, 14.2, 16.3, 18.2, 17.1, 20.2, 22.4]

[Link](X, Y)
[Link]("day") # x-axis label
[Link]("temperature") # y-axis label
[Link]()

Robust Physical AI School of Energy

& Robotics Lab. Engineering
 Matplotlib
Matplotlib
: one of the most widely used Python plotting libraries
> Line plot, box plot, bar charts, Pie charts, 3D plot
> 1D data, 2D image, 3D data, …
import [Link] as plt

X = [ "Mon", "Tue", "Wed", "Thur", "Fri", "Sat", "Sun" ]

Y1 = [15.6, 14.2, 16.3, 18.2, 17.1, 20.2, 22.4]
Y2 = [20.1, 23.1, 23.8, 25.9, 23.4, 25.1, 26.3]

[Link](X, Y1, label="Seoul")

[Link](X, Y2, label="Busan")
[Link]("day")
[Link]("temperature")
[Link](loc="upper left")
[Link]("Temperatures of Cities")
[Link]()
Robust Physical AI School of Energy
& Robotics Lab. Engineering
 Matplotlib
Matplotlib
: one of the most widely used Python plotting libraries
> Line plot, box plot, bar charts, Pie charts, 3D plot
> 1D data, 2D image, 3D data, …
import [Link] as plt

Y1 = [15.6, 14.2, 16.3, 18.2, 17.1, 20.2, 22.4]

Y2 = [20.1, 23.1, 23.8, 25.9, 23.4, 25.1, 26.3]

[Link]([Y1, Y2]
[Link]([1, 2], [“City A”, “City B”])
[Link]("temperature")
[Link]("Temperatures of Cities")
[Link]()

Robust Physical AI School of Energy

& Robotics Lab. Engineering
 Matplotlib
import [Link] as plt

[Link]([15.6, 14.2, 16.3, 18.2, 17.1, 20.2, 22.4], "sm")

[Link]()

X = [ "Mon", "Tue", "Wed", "Thur", "Fri", "Sat", "Sun" ]

Y = [15.6, 14.2, 16.3, 18.2, 17.1, 20.2, 22.4]
[Link](X, Y)
[Link]()

Y = [38, 22, 15, 25]

labels = ["Apples", "Pear", "Strawberry", "Cherries"]
explode = [0.1, 0, 0, 0]

[Link](Y, labels = labels, explode = explode)

[Link]()

Robust Physical AI School of Energy

& Robotics Lab. Engineering
 Lab 1. Matplotlib
1. [Performance comparison graph] Let’s say you have trained three different AI models:
• Model 1: accuracy 64.3 %
• Model 2: accuracy 79.3 %
• Model 3: accuracy 95.3 %
Draw three types of graphs: a bar graph, line plot, and pie chart, and discuss which of these graphs the
most effectively shows the performance difference
Example input/output:

Robust Physical AI School of Energy

& Robotics Lab. Engineering
Word Cloud
Word Cloud
: Generates visual “word clouds” from text frequency
: Larger font size = more frequent word
: Useful for summarizing text, exploring NLP datasets, and analyzing news or social
media

Robust Physical AI School of Energy

& Robotics Lab. Engineering
 Word Cloud
import [Link] as plt
from wordcloud import WordCloud

text=""
with open("[Link]", "r", encoding="utf-8") as f:
lines = [Link]()
for line in lines:
text += line
text = "Python text data visualization wordcloud"

wc = WordCloud(width=600, height=400)

[Link](text)
wc.to_file("[Link]")

[Link](figsize=(30, 10))
[Link](wc)
[Link]()

Robust Physical AI School of Energy

& Robotics Lab. Engineering
NumPy and OpenCV

NumPy (Numerical Python) OpenCV

: Efficient data representation : Open-source computer vision library
(ndarray) for scientific computing (image, video)
Package: [Link] Package: [Link]
Source: [Link] Source: [Link]
Document: [Link] Document:
[Link]

Robust Physical AI School of Energy

& Robotics Lab. Engineering
NumPy and OpenCV
$ pip install numpy opencv-python
> On your window CMD, powershell, VScode terminal

Robust Physical AI School of Energy

& Robotics Lab. Engineering
NumPy
NumPy (Numerical Python)
: The fundamental package for scientific computing, numerical operation, linear
algebra, matrix operations.
: Provides ndarray, a fast and memory-efficient n-dimensional array
: Backbone lineary used by PyTorch, TensorFlow, OpenCV, Pandas, SciPy, etc
: Much faster than Python lists

Robust Physical AI School of Energy

& Robotics Lab. Engineering
NumPy
import numpy as np
ndarray
# (1) creating arrays
• In NumPy, a Fundamental data structure x_1d = [Link]([1, 2, 3.2]) #1D array
used to represent multi-dimensional arrays x_2d = [Link]([[1,2],[3,4]]) # 2D
x_3d = [Link](3, 256, 256) # 3D
• Scalar (0D), Vector (1D), Matrix (2D), x_4d = [Link](10, 3, 256, 256) # 4D
x_5d = [Link](4,3,3,5,5) # 5D
higher-dimensional array (≥ 3D)
• Attribute: # (2) printing ndarray properties
print(x_2d.shape, x_2d.dtype,
• shape : dimensions x_2d.size, , x_2d.ndim)
# (2,2), float64, 4, 2
• dtype : data type
# more...
• size : total number of elements dir(x_2d)
• ndim: rank (# of axes)

Robust Physical AI School of Energy

34
& Robotics Lab. Engineering
 Motor input voltage
Tensor PWM signal
(T × Voltage)
• Scalar (0D) [12V, 6V, 3V, 12V, …]
: e.g., single data point, such as
temperature (25°C), voltage value, … Time series representation
(T × Power (kWh))
[100, 200, 210, …]
• Vector (1D) Power consumption,
: e.g., Time-series scalar data generation, loss, …
(audio, electric signal, …), feature, …
Audio waveform
: 1D amplitudes
(T × 16,000)
• Matrix (2D)
[-0.1, 1.0, 0.7, …]
: e.g., Time-series vector data, Gray-scale Assuming sampling at 16KHz
…
image, …
* Other electric waves also have similar form [-0.3, 0.1, 0.4, …]
(Bluetooth, 5G, communication, medical equipment)

Robust Physical AI School of Energy

35
& Robotics Lab. Engineering
 Audio spectrogram
Tensor : 2D representation
(frequency × Time)
• Matrix (2D)
: e.g., Time-series vector data, Gray-scale Video
RGB image
image, … 127 0 …
127 0 …
127 0 …
64 2 …
64 2 …
64 2 …
• Tensor (3D Data) …
…
…
…
…
…

: e.g., RGB-scale image, mesh, … … … …

RGB values × resolution Time × RGB values × resolution

3×3840×2160 T×3×3840×2160
• Tensor (4D Data) LiDAR
# of measurement × [X, Y, Z, Intensity]
: e.g., RGB video, Lidar point, … N×4
[0, 0, 0, 128]
[5, 3, 64, 128]
[13, 77, 0, 128]
• Higher-dimensional array (≥ 5D) …
: e.g., Lidar point cloud, RADAR, … LiDAR sensor

Robust Physical AI School of Energy

36
& Robotics Lab. Engineering
NumPy – Slicing & Indexing
• Indexing & slicing
# (3) indexing, slicing, Boolean masking 0 1 2
import numpy as np
array2d = [Link]([[10, 20, 30], 0 10 20 30
[40, 50, 60],
[70, 80, 90]])
1 40 50 60

# (3-1) indexing: (0,1)

70 80 90
print(f’array2d[0, 1] = {array2d[0, 1]}') # 20 2
# (3-2) row slicing: from row 1 to the end
print(f’array2d[1:, :] =\n {array2d[1:, :]}’)

# (3-3) Boolean masking

mask = array2d > 50
print(f'mask:\n {mask}')
print(f’bigger than 50: {tensor2d[mask]}')

Robust Physical AI School of Energy

37
& Robotics Lab. Engineering
NumPy – Linear Algebra Operation
• Linear Algebra Operations
# (4) Linear Algebra Operations
import numpy as np

A2d = [Link]([[1,2],[3,4]])
B2d = [Link]([[5,6],[7,8]])

print([Link](A2d, B2d)) # matrix multiplication

print([Link](A2d)) # matrix transpose
print([Link](A2d)) # matrix inverse
print([Link](A2d)) # eigenvalues

# (5) other ndarray methods

print([Link]()); print([Link]()); print([Link]());
print([Link]()); print([Link]()); print([Link]())

Robust Physical AI School of Energy

38
& Robotics Lab. Engineering
NumPy – list to NumPy
• Converting list to numpy array
# (6) List to NumPy
import numpy as np
import [Link] as plt

ftemp_lst = [63, 73, 80, 86, 84, 78, 66, 54, 45, 63]
ftemp_ary = [Link](ftemp_lst)

print(ftemp_lst) # python list

print(ftemp_ary) # numpy ndarray

ctemp_ary = (ftemp_ary-32.)*5/9
[Link](ctemp_ary)
[Link]("index")
[Link]("temperature")
[Link]()

Robust Physical AI School of Energy

39
& Robotics Lab. Engineering
 NumPy – list to NumPy
• NumPy creating with NumPy methods (arrange, linspace, rand)
# (7) Array with arrange, linsapce, rand
import numpy as np
import [Link] as plt

arange_array = [Link](0, 10, 2) # (start, stop, step), 0, 2, 4, 6, 8

linspace_array = [Link](0, 1, 5) # (start, stop, num), 0, 0.25, 0.5, 0.75, 1
[Link](arange_array)
[Link]()

X = [Link](-2*[Link], 2*[Link], 100) # [-2𝜋, 2𝜋], 100 samples

Y1 = [Link](X) # Y = sin(X)
Y2 = 3*[Link](X) # Y = 3*sin(X)

[Link](X,Y1,X,Y2)
[Link]()

Robust Physical AI School of Energy

40
& Robotics Lab. Engineering
 NumPy – list to NumPy
• NumPy creating with NumPy methods (arrange, linspace, rand)
# (7) Array with arrange, linsapce, rand
import numpy as np
import [Link] as plt

rand_1d_array = [Link](5) # uniform distribution within [0,1]

randn_1d_array = [Link](5) # gaussian normal dist(mean:0, std:1)
rand_2d_array = [Link](5,5)# 2d array
randn_1d_array = [Link](1, 2, 100) # gaussian normal dist(mean:1, std:2)

mu, sigma = 10, 2

Y1 = [Link](10000)
Y2 = [Link](10000)
Y3 = mu + sigma*[Link](10000)
[Link](Y1,bins=20); [Link]()
[Link](Y2,bins=20); [Link](Y3,bins=20); [Link]()

Robust Physical AI School of Energy

41
& Robotics Lab. Engineering
Lab 2. NumPy
1. [Draw graph] Draw graph 1(𝑌 = 1, 𝑌 = 𝑋, 𝑌 = 𝑋 2 ) and graph 2 (𝑌 = 𝑋 + 𝑁𝑜𝑖𝑠𝑒), Noise is
sampled from guassian normal distributions

Example input/output:

Robust Physical AI School of Energy

42
& Robotics Lab. Engineering
OpenCV
OpenCV (Open Source Computer Vision Library)
: A powerful library for image processing, computer vision, feature extraction,
object detection, video processing
: written in C/C++, with Python bindings
: widely used in AI and Robotics

Image blending

Template Matching Object tracking Classification, detection, segmentation

Robust Physical AI School of Energy
& Robotics Lab. Engineering
OpenCV import cv2
import [Link] as plt

Reading, Showing, Writing Image # (1) reading and writing images

# OpenCV loads images as BGR, not RGB
• OpenCV utilizes NumPy’s ndarray to img = [Link]("opencv_logo.png") # BGR
represent images or videos [Link](“output_img.jpg”, img)

• Attribute: # (2) displaying images

[Link]("Image", img)# need BGR format
• shape : dimensions [Link](0)
(height, width, channels) [Link]()

• dtype : data type [Link](img) # need RGB format

(uint8, 0 - 255) [Link]()

• size : total number of elements # (3) Image properties

print([Link], [Link], [Link],
• ndim: rank (# of axes) [Link])
# (739,600,3), uint8, 1330200, 3

Robust Physical AI School of Energy

46
& Robotics Lab. Engineering
 OpenCV
• Converting Color Space
import cv2
import [Link] as plt
import numpy as np

# (4) Converting Color Space

img = [Link]("opencv_logo.png") # BGR
rgb = [Link](img, cv2.COLOR_BGR2RGB) # RGB
gray = [Link](img, cv2.COLOR_BGR2GRAY) # GRAY

[Link](1,3,1); [Link](img); [Link]("off");

[Link](1,3,2); [Link](rgb); [Link]("off");
[Link](1,3,3); [Link]([Link](gray[:,:,None],3,axis=2));[Link]("off");
[Link]("output_img.jpg")
[Link]()

Robust Physical AI School of Energy

& Robotics Lab. Engineering
 OpenCV
• Image Resizing & Cropping
import cv2
import [Link] as plt
import numpy as np

# (5) Image Resizing & Cropping

img = [Link]("kentech_logo.png") # BGR, (110, 459, 3)
resized = [Link](img, (300,300)) # BGR, (300, 300, 3)
cropped = img[70:110, :, :] # BGR, (40, 459, 3)

[Link]("Image", img); [Link](1000)

[Link]("Image", resized); [Link](1000)
[Link]("Image", cropped); [Link](1000)
[Link]()

Robust Physical AI School of Energy

& Robotics Lab. Engineering
 OpenCV
• Image Blending
import cv2
import [Link] as plt
import numpy as np

# (6) Image Blending

img = [Link]("kentech_logo.png") # BGR, (110, 459, 3)
resized = [Link](img, (300,300)) # BGR, (300, 300, 3)

img2 = [Link]("opencv_logo.png") # BGR, (739, 600, 3)

resized2 = [Link](img2, (300,300)) # BGR, (300, 300, 3)

img3 = (0.5*resized + 0.5*resized2)).astype(np.uint8)

[Link]("Image", img3)
[Link](0)
[Link]()
Robust Physical AI School of Energy
& Robotics Lab. Engineering
 OpenCV
• Edge Detection (Canny) & Blurring
import cv2
import [Link] as plt
import numpy as np

# (7) Edge detection & Blurring

img = [Link]("kentech_logo.png",cv2.IMREAD_GRAYSCALE) # gray, (110, 459)
edges = [Link](img, 100, 200) # lower/upper threshold
blur = [Link](img, (5,5), 0) # kernel size (5x5), sigma 0

[Link]("Image", img); [Link](1000)

[Link]("Image", edges); [Link](1000)
[Link]("Image", blur); [Link](1000)
[Link]()

Robust Physical AI School of Energy

& Robotics Lab. Engineering
 OpenCV
• Drawing on Images
import cv2
import [Link] as plt
import numpy as np

# (8) Drawing on Images

img = [Link]("opencv_logo.png")
img2 = [Link](img, (100,100), 30, (0,255,0), 3) # center,radius,color,thick
img3 = [Link](img, (50,50), (150,150), (255,0,0), 2)
img4 = [Link](img, "Hello", (50,200), cv2.FONT_HERSHEY_SIMPLEX, 1, (0,0,0),
2)

[Link]([Link](img, cv2.COLOR_BGR2RGB))
[Link]("off")
[Link]()

Robust Physical AI School of Energy

& Robotics Lab. Engineering
 OpenCV
• Color distribution analysis
import cv2
import [Link] as plt
# (9) Color distribution analysis
img = [Link](＂kentech_logo.png") # BGR
img_B = img[:,:,0] # B
img_G = img[:,:,1] # G
img_R = img[:,:,2] # R
B, G, R = [Link](img)

[Link](figsize=(10,5))
[Link]([Link]([R], [0], None, [256], [0,256]), color='red')
[Link]([Link]([G], [0], None, [256], [0,256]), color='green')
[Link]([Link]([B], [0], None, [256], [0,256]), color='blue')
[Link]("RGB Color Histogram“); [Link]("Pixel Intensity");
[Link]("Frequency"); [Link](["Red","Green","Blue"])
[Link]()
Robust Physical AI School of Energy
& Robotics Lab. Engineering
 OpenCV
• Webcam
import cv2

# (10) Basic webcam (video) processing

cap = [Link](0)
While True:
ret, frame = [Link]() # Capture a frame
if not ret:
break
[Link]("Webcam", frame) # Display the frame

# Press 'q' to exit

if [Link](1) & 0xFF == ord('q'):
break

[Link]() # Release the camera

[Link]()
Robust Physical AI School of Energy
& Robotics Lab. Engineering
Lab 3. OpenCV
1. [Image Composition] Download two random images (one: background, one: foreground image with
green chroma key) and composite them together.

Example input/output:

Robust Physical AI School of Energy

54
& Robotics Lab. Engineering
Summary
[Python Library]
• [Recap] Important thing is a logical thinking!
• Python Library
• Matplotlib
: Data visualization (histogram, heat map, pie chart, graph, …)
• Word Cloud
: Generates visual “word clouds” from text frequency
• Pandas
: Provides DataFrame and Series structures for CSV, Excel, and JSON data
• Numpy
: Fundamental package for scientific computing, numerical operation, linear algebra, matrix
operations.
• OpenCV
: A powerful library for image processing, computer vision, feature extraction, object detection,
video processing
Robust Physical AI School of Energy
55
& Robotics Lab. Engineering
 Q&A

Robust Physical AI School of Energy

56
& Robotics Lab. Engineering