Developing Image Processing and Computer

Vision applications
Artificial Intelligence

Dr. Usman Ghani

INDEX

▰ Introduction to OpenCV
▰ Image Basics in OpenCV
▰ Basic functions of OpenCV
▰ Constructing Basic Shapes in OpenCV
▰ Image filtering
▰ Morphological transformations
▰ Processing Images with OpenCV
▰ Working with videos

2
Introduction to
OpenCV

3
How does computer vision works?

4
 Applications of OpenCV

Human Pose Detection Detection and Tracking

• e.g. Tracking Cars or Objects
• e.g. AR clothes fitting

Medical Imaging Camera Calibration

• e.g. camera calibrations toolkits
• e.g. AI medical diagnosis

Manufacturing Production
Autonomous vehicles Line
• e.g. Self Driving Cars • Labeling and tracing products

5
Computer Vision Toolkits

Endrov
VTK VXL

PIL BootCV

Computer
PCL Vision VLFeat
toolkits

6
Introduction to the OpenCV library

▰ Open source Computer Vision Library

▰ Officially launched in 1999 at Intel Research
▰ Available for C, C++, and Python
▰ free for both academic and commercial
▰ Cross Platform
▻ Supports Windows, Linux, Mac OS , iOS, Android
▰ http://opencv.org/

7
Installing OpenCV, Python, and other packages

▰ Installation of required ▰ Installation Commands:

libraries:  pip install opencv-python
 Install opencv-python  pip install matplotlib
 Install matplotlib  pip install scikit-image
 Install scikit-image  pip install numpy
 Install numpy  pip install pandas
 Install pandas

8
Image Basics in
OpenCV

9
Concepts of pixels, colors, channels, images,
and color spaces

Pixels Colors Images Color

channels
spaces

10
 Accessing and manipulating pixels in OpenCV

Access pixel values

Access only blue pixel

Modify the pixel values

11
 Get image shape

• type of image data; number of pixels; etc.

• print( img.shape )

12
Basic functions of
OpenCV

13
 Read an Image

Syntax: cv2.imread(image, flag)

14
https://miro.medium.com/max/1400/1*IHHLvhkhx2xHloOb3-wXwA.jpeg
Display an image

Syntax: cv2.imshow(window_name, image)

Output
15
Save image to disk

Syntax: cv2.imwrite(filename, image)

Image saved in drive

16
 Access Image properties

Syntax:
• img.shape
• img.dtype
• img.size

17
Change Color Spaces

RGB → Gray
cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)

18
Resize image

Syntax:cv2.resize(src, dsize[, dst[, fx[, fy[, interpolation ]]]])

Original Image Resized Image

19
Constructing Basic
Shapes in OpenCV

20
Draw a rectangle on image

Syntax: cv2.rectangle(image, p0, p1, color,

thickness)

Output
21
Draw a circle on image

Syntax: cv2.circle(image,
center_coordinates, radius, color,
thickness)

Output
22
Draw a line on image

Syntax: cv2. line(image, p0, p1, color,

thickness)

Output
23
Draw an ellipse on image

Syntax: cv2.ellipse(img, center, axes, angle,

a0, a1, color, thickness)

Output
24
Put text on image

Syntax: cv2.putText(image, text, org, font,

fontScale, color[, thickness[, lineType[,
bottomLeftOrigin]]])

Output
25
Image filtering

26
Sharpen the image

Syntax: cv2.filter2D(src, depth, k_sharped,

kernel)

Output
27
Blur an image

Syntax: cv2.blur(image, ksize)

Output
28
Apply Gaussian Filter on an image

Syntax: cv2.GaussianBlur(src, ksize, sigmaX[,

dst[, sigmaY[, borderType =
BORDER_DEFAULT]]] )

Output
29
Apply Median Filter on an image

Syntax: cv2.medianBlur(src, ksize)

Output
30
Apply an identity Filter on an image

k_filter = np.array([[1,0,0], [1,1,0], [0,0,1]])

Syntax: cv2.filter2D(img, -1, k_filter)

Output
31
Processing Image
Operations with OpenCV

32
Perform Thresholding on an image

Syntax: cv2.threshold(image, thresh, maxval,

type, dst)

Output
33
Detect Edges of objects in an image

Syntax: cv2.Canny('/path/to/img', minVal,

maxVal, apertureSize, L2gradient)

Output
34
Detect Contours in an image

#find Contours
cv2.FindContours(image,mode=
CV_RETR_LIST,method=CV_CHAIN_APPROX_
SIMPLE)
#draw contours
cv2.DrawContours(img, contours,
contourIdx, colour, thickness)

Output
35
Setting Region of Image

Syntax: cv2.selectROI(windowName, img,

showCrosshair, fromCenter)

Output
36
 Blending Images

Blend Pictures
Syntax: cv2.addWeighted(img1, alpha, img2, beta, gamma)

Pic 1 Pic 2

37
 Crop an image

Syntax: crop_img = img[y:y+h, x:x+w]

Original Image Cropped Image

38
 Splitting and merging images

Syntax: Merged an image

• cv2.split()
• cv2.merge()

Split an image

39
Draw Histogram of an image

Syntax: cv2.calcHist(images, channels, mask,

histSize, ranges)

Original Grayscale Image

40
Apply Log Transformation on an image
# Apply log transform.
c = 255/(np.log(1 + np.max(img)))
log_transformed = c * np.log(1 + img)
abc= np.array(log_transformed, dtype = np.uint8)

Input Image Apply Log Transformation

41
 Apply Power Law Transformation on an image

Original Image Gamma = 2.2 Gamma = 0.1

Pow_law_trans = np.array(255*(img / 255) ** gamma, dtype = 'uint8')

42
Morphological
Transformations

43
Dilation Operation

Syntax: cv2. dilate(img, kernel, iterations=1)

Input Image Dilation of Image

44
Erosion Operation

Syntax: cv2.erode(img, kernel, iterations=1)

Input Image Erosion of Image

45
Opening Operation

Syntax: cv2.morphologyEx(img,
cv2.MORPH_OPEN, kernel)

Input Image Opening operation on Image

46
Closing Operation

Syntax: cv2.morphologyEx(img,
cv2.MORPH_CLOSE, kernel)

Input Image Closing operation on Image

47
Top Hat Operation

Syntax: cv2.morphologyEx(img,
cv2.MORPH_TOPHAT, kernel)

Input Image Top Hat Operation

48
Black Hat Operation

Syntax: cv2.morphologyEx(img,
cv2.MORPH_BLACKHAT, kernel)

Input Image Black Hat Operation

49
Morphological gradient operation

Syntax: cv2.morphologyEx(img,
cv2.MORPH_GRADIENT, kernel))

Input Image Morphological gradient operation

50
Perform Negative of an Image
# Subtract the img from max value
(calculated from dtype)
img_neg = 255 - img

Input Image Perform Negative Operation

51
 THANKS!
Prof Dr. Muhammad Usman Ghani Khan
Director
NCAI UET Lahore
[email protected]

52