Index in position 1 exceeds array bounds.

As a task I have to implement a RANSAC algorithm. There is an array called correspondences which is 4xn, where the first 2 rows are x1 and y1 and the last two rows are x2 and y2. These are coordinates for pixels. The task here is to seperate the coordinates into their own homogenious variables. Before that one must provide some optional variables for the function. I used an Input Parser for that and it seems to work. This is what I did:

function [epsilon, p, tolerance, x1_pixel, x2_pixel] = F_ransac(correspondences, varargin)

% This function implements the RANSAC algorithm to determine

% robust corresponding image points

p1 = inputParser;

addOptional(p1, 'epsilon', 0.5, @(x)validateattributes(x, {'numeric'}, {'<',1}, {'>',0}));

addOptional(p1, 'p', 0.5, @(x)validateattributes(x, {'numeric'}, {'<',1}, {'>',0}));

addOptional(p1, 'tolerance', 0.01, @(x)validateattributes(x, {'numeric'}));

parse(p1);

epsilon = p1.Results.epsilon;

p = p1.Results.p;

tolerance = p1.Results.tolerance;

x1_pixel = ones(3,size(correspondences,2));

x1_pixel(1:2,:) = correspondences(1:2,:);

x2_pixel = ones(3,size(correspondences,2));

x2_pixel(1:2,:) = correspondences(3:4,:);

end

This is constantly giving me the error:

Index in position 1 exceeds array bounds.

Error in F_ransac (line 16)

x1_pixel(1:2,:) = correspondences(1:2,:);

Why is it giving me this error? What could I do to avoid it?

[VBBV]

correspondences = rand(3,5); % less than 4 
F_ransac(correspondences,3)

Index in position 1 exceeds array bounds. Index must not exceed 3.

Error in solution>F_ransac (line 19)
x2_pixel(1:2,:) = correspondences(3:4,:);

function [epsilon, p, tolerance, x1_pixel, x2_pixel] = F_ransac(correspondences, varargin)
    % This function implements the RANSAC algorithm to determine 
    % robust corresponding image points
    
    p1 = inputParser;
    addOptional(p1, 'epsilon', 0.5, @(x)validateattributes(x, {'numeric'}, {'<',1}, {'>',0}));
    addOptional(p1, 'p', 0.5, @(x)validateattributes(x, {'numeric'}, {'<',1}, {'>',0}));
    addOptional(p1, 'tolerance', 0.01, @(x)validateattributes(x, {'numeric'}));
    parse(p1);
    epsilon = p1.Results.epsilon;
    p = p1.Results.p;
    tolerance = p1.Results.tolerance;
    x1_pixel = ones(3,size(correspondences,2));
    x1_pixel(1:2,:) = correspondences(1:2,:);
    x2_pixel = ones(3,size(correspondences,2));
    x2_pixel(1:2,:) = correspondences(3:4,:);
end 

May be you are calling function with input arguments of incorrect size

Answer 1

[Chunru]

It seems that your code works well. Can you show how the error occurs?

x = randn(4, 5)

x = 4×5

1.7007 1.0577 1.2506 0.4107 1.0085 -0.3642 0.0145 0.4001 -0.3132 0.0327 0.0062 -0.1212 0.1764 -1.7733 -1.0230 -0.3822 -0.4836 0.4640 -0.9320 -1.5065

F_ransac(x)

ans = 0.5000

function [epsilon, p, tolerance, x1_pixel, x2_pixel] = F_ransac(correspondences, varargin)
    % This function implements the RANSAC algorithm to determine 
    % robust corresponding image points
    
    p1 = inputParser;
    addOptional(p1, 'epsilon', 0.5, @(x)validateattributes(x, {'numeric'}, {'<',1}, {'>',0}));
    addOptional(p1, 'p', 0.5, @(x)validateattributes(x, {'numeric'}, {'<',1}, {'>',0}));
    addOptional(p1, 'tolerance', 0.01, @(x)validateattributes(x, {'numeric'}));
    parse(p1);
    epsilon = p1.Results.epsilon;
    p = p1.Results.p;
    tolerance = p1.Results.tolerance;
    x1_pixel = ones(3,size(correspondences,2));
    x1_pixel(1:2,:) = correspondences(1:2,:);
    x2_pixel = ones(3,size(correspondences,2));
    x2_pixel(1:2,:) = correspondences(3:4,:);
end 

Answer 2

[Matis Tartie]

Yes I see. I just noticed there is something wrong with the correspondences array. Thank you for your time