use detectORBFeatures at r2018a

hi, i want to use detectORBFeatures function

but my matlab version is R2018a, and by my searched result, detectORBFeatures is developed at matlab 2019 version, so i can't use that function at my matlab environment.

Is there any way i can use detectORBFeatures at my matlab environment (r2018a)?

Or, which functions related to feature based alignment that i can use at my matlab version?

thanks a lot for all your responses :))

Answer 1

[Angelo Yeo]

(1) You can use alternative detectors that are introduced before R2019a. For example, these detectors find corner features like ORB does.

detectFASTFeatures

detectMinEigenFeatures

detectHarrisFeatures

(2) However, the feature type should not be the only criteria when you choose detectors. See the doc below for further explanation.

Local Feature Detection and Extraction - MATLAB & Simulink (mathworks.com)

(3) Below is an example to compare the detection algorithms I mentioned earlier with ORB. (Note: This example is taken from detectORBFeatures documentation.)

%% 
% Read an image into the workspace.

I = imread('businessCard.png');
%% 
% Convert the image into a grayscale image.

I = im2gray(I);
%% 
% Display the grayscale image.

figure
imshow(I)

%% 
% Detect and store keypoints using ORB, FAST ,MinEigen, Harris detectors

pointsORB = detectORBFeatures(I); % Introduced in R2019a
pointsFAST = detectFASTFeatures(I); % Introduced in R2013a
pointsMinEigen = detectMinEigenFeatures(I); % Introduced in R2013a
pointsHarris = detectHarrisFeatures(I); % Introduced in R2013a

%% 
figure
subplot(2,2,1)
imshow(I)
hold on
plot(pointsORB.selectStrongest(100),"ShowScale", false)
hold off;
title("ORB")

subplot(2,2,2)
imshow(I)
hold on
plot(pointsFAST.selectStrongest(100))
hold off;
title("FAST")

subplot(2,2,3)
imshow(I)
hold on
plot(pointsMinEigen.selectStrongest(100))
hold off;
title("MinEigen")

subplot(2,2,4)
imshow(I)
hold on
plot(pointsHarris.selectStrongest(100))
hold off;
title("Harris")