4 x 4 Mask to Identify Regions on Map with Highest Value

Good afternoon everyone!

I have a 360 x 576 matrix that represents a map of the world (latitude and longitude, respectively). Each of the elements in the matrix has a value between 0 and 100 that represents the amount of precipitation present at that geographic location. I would like to create a mask that checks every 4 rows by 4 columns to identify those regions on the "map" with the highest concentration of values. Essentially, I am looking for those regions where intense precipitation concentrates so that I may further inspect these geographic locations. Here, I think finding the average of the 4 x 4 mask might be helpful, but I don't know how to apply the averaging method to the entire matrix without overlapping. Also, I'm unsure if I can "highlight" the top five or so "4 x 4 boxes" with the highest values for the entire matrix. Any suggestions? I'm actively working on the problem but would appreciate any ideas. Thanks!

Michelle

Answer 1

[Image Analyst]

Try this:

% Demo to find 4x4 regions with the highest sum value.
clc;    % Clear the command window.
fprintf('Beginning to run %s.m ...\n', mfilename);
close all;  % Close all figures (except those of imtool.)
imtool close all;  % Close all imtool figures.
clear;  % Erase all existing variables.
workspace;  % Make sure the workspace panel is showing
fontSize = 15;

% Get temperature array with temps between 20 and 30 degrees celsius.
grayImage = 20 + 10 * rand(360, 576);
subplot(2, 1, 1);
imshow(grayImage, [])
impixelinfo; % Let user see RGB values as they mouse around.
colorbar
axis('on', 'image');
title('Original Image', 'FontSize', fontSize);
% Maximize the window to make it easier to draw.
g = gcf;
g.WindowState = 'maximized'
fprintf('Done running %s.m ...\n', mfilename);

% Get kernel
kernel = ones(4, 4);
sumImage = conv2(grayImage, kernel, 'same');
subplot(2, 1, 2);
imshow(sumImage, []);
axis('on', 'image');
title('Sum Image', 'FontSize', fontSize);

% Find 5 highest sums
sortedValues = sort(sumImage(:), 'descend');
[highRows, highColumns] = find(sumImage >= sortedValues(5))
% Place red squares around those areas.
hold on;
MarkerSize = 30;
plot(highColumns, highRows, 'rs', 'MarkerSize', MarkerSize, 'LineWidth', 3);

[Michelle De Luna]

Thank you for your response! I truly appreciate the help.

Answer 2

[jonas]

Using imresize from the image processing TB

%non-overlapping (use conv2 method 'valid' for overlapping) 
A = imresize(Z,0.5,'box');
%get indices for max id
[val, id] = max(A,[],[1,2],'linear');
[x,y] = ind2sub(size(A),id)*2;

The indices are multiplied by a factor 2 to get indices in the original matrix.

[Michelle De Luna]

Thank you for your help, Jonas! This was very insightful! :)