20bcs7575CV Lab 2.1

Uploaded by

Ashu Rajput

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

18 views4 pages

20bcs7575CV Lab 2.1

Uploaded by

Ashu Rajput

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 4

CourseName: ComputerVision Lab CourseCode: CSP-422

Experiment- 2.1

Student Name: Aashish singh chauhan UID: 20bsc7557

Branch: CSE Section/Group: 20BCS-8A
Semester: 7th Subject Code:CSP-422
Subject Name: CV Lab

Aim: To evaluate the efficacy of human guided control point selection for image alignment.
SoftwareRequired: MATLAB
Description:
➢ Human-guided control point selection: Human-guided control point selection is a technique used in
image alignment tasks where a human operator manually selects a set of control points to guide the
alignment process. These control points are typically landmarks or distinctive features in the images that
can be easily identified by a human observer.

Human-guided control point selection allows the human operator to leverage their visual perception
and domain knowledge to guide the alignment process. It is particularly useful when dealing with
challenging alignment scenarios or when accurate alignment is critical. However, it can be
timeconsuming and subjective, as the quality of the alignment heavily relies on the operator's expertise.
Therefore, automated techniques such as feature-based matching or deep learning-based approaches
can be employed to reduce human involvement or assist in the control point selection process.

Algorithm:
1. Load the source and target images.
2. Display the source and target images to the user.
3. Prompt the user to select corresponding control points on both images.
4. Store the selected control points for alignment.
5. Perform image alignment using the control points.
6. Display the aligned image to the user.
7. Evaluate the alignment quality using metrics such as mean square error or structural similarity
index.
8. Calculate the efficacy of human-guided control point selection by comparing the alignment
results with and without user intervention.
9. Output the evaluation results and efficacy metrics.

Name:AbhaySharma UID:20BCS7575
CourseName: ComputerVision Lab CourseCode: CSP-422

Implementation:
➢ Human-guided controlpoint selection:

movingImage = imread("Orion1.png");
fixedImage = imread("Orion2.png");
cpselect(movingImage, fixedImage);

tform = estgeotform2d(movingPoints, fixedPoints, "similarity"); movingImageT =

imwarp(movingImage, tform, "OutputView", imref2d(size(fixedImage)));
montage({fixedImage, movingImageT})

% Calculate alignment evaluation metrics

% You can use any suitable metric to compare alignedImageManual and alignedImageAuto
% For example, mean squared error (MSE), structural similarity index (SSIM), etc.

mse = immse(fixedImage, movingImageT); ssimValue

= ssim(fixedImage, movingImageT);

fprintf('MSE between ground truth and automated alignment: %.4f\n', mse); fprintf('SSIM
between ground truth and automated alignment: %.4f\n', ssimValue);
Output:
➢ Human-guided control point selection (Accuracy and Result):

Name:AbhaySharma UID:20BCS7575
CourseName: ComputerVision Lab CourseCode: CSP-422