SRA-Tasks

Image Processing Tasks

Tables of Contents

• About
• Task 1
• Task 2
• Task 3
• Task 4
• Task 5
• Task 6

About

All the tasks are performed using libraries such as Numpy and Pillow, without using other inbuilt functions from libraries like Open-CV.

Task 1- Image Rotation :

Rotating the given image by various angles without the use of inbuilt rotate functions of numpy, PIL or OpenCV.

Rotation Matrix:

Original image	Result After Rotation(45 degrees)

Task 2 - Applying Kernels :

Blurring An Image

Blurring the image with 5x5 kernels. Box-Blur and Gaussian Blur Kernels are used.

Original image	Box Blur	Gaussian Blur

Original image	Box Blur	Gaussian Blur

Sharpening Image

Sharpening the image with 5x5 kernel.

Original Image	Sharpened Image

Original Image	Sharpened Image

Task 3 - Edge Detection :

Edge Detection done using 3x3 kernels.

Original Image	Vertical Edge Detection	Horizontal Edge Detection
Sobel Edge Detection	Canny Edge Detection

Original Image	Vertical Edge Detection	Horizontal Edge Detection
Sobel Edge Detection	Canny Edge Detection

Task 4 - Morphology :

Erosion and Dilation

Original Image	Erosion	Dilation

Task 5 - Masking :

Input Image	Blue Ball Detection

Task 6 - ROI

Region of Interest extraction

Input Image	Output Image