Multi-Dimensional Arrays

Return to List of Lectures

Previous Lecture

Misc

One-Button Studio

New access procedure: sign up for time via the Google calendar accessed from cit.geneseo.edu > media > video guide.

This is important -- it’s the only way to ensure you aren’t barging in on someone else’s recording!

Calculus Bee

Wednesday, April 19, 2:45 PM, Newton 204

Colloquium

“Precision Medicine in the Age of ‘Big Data’: Leveraging Machine Learning and Genomics for Drug Discoveries”

Katie Gayvert, Weill Medical College (Cornell) and Geneseo math alumna, class of 2012.

Monday, April 17, 2:30, Newton 203.

Extra credit for a write-up.

Questions?

Indexing Multidimensional Arrays

Sections 8.5 - 8.7

Indexing

Pick out the top row of the picture of me on Mars. How about the left column? How about the top half of the image?

Relevant ideas or questions from the reading:

• You can use linear indexing to pick out elements by integer position.
• Or you can pull out whole rows with the single colon index.
• You can replace a row or other elements using the single colon too.
• You can use logical indexing with arrays of the same dimensions.

Think of a 3-dimensional array as made up of rows, columns, and planes:



Solutions:

• The first row of the image (and imshow really does display it as a single row of pixels):

>> row1 = baldwinMars( 1, :, : );
>> imshow( row1 )

• The first column of the image (again, displayed by imshow as a tall column of color):

>> col1 = baldwinMars( :, 1, : );
>> figure(3);imshow( col1 );

• The top half (roughly) of the image:

>> half = baldwinMars( 1:190, :, : );
>> imshow( half )

Comments:

• The key things here are how to index “slices” of an array. In particular colon to select all possible indices in a certain dimension, or a colon-separated range to select some of the indices in a dimension.
• Single numbers also index a single row/column/etc.

Changing Things

Make the top half of the picture of me on Mars pure red. Notice that you have to set the red channel to the maximum color value and the green and blue channels to the minimum.

>> baldwinMars( 1:190, :, 1 ) = 255 * ones( 190, 516 );
>> imshow( baldwinMars )
>> baldwinMars( 1:190, :, 2:3 ) = zeros( 190, 516, 2 );
>> imshow( baldwinMars )

Comments:

• This demonstrated how you can use indexing and assignment to pick out a part of an array you want to change.

Next

Flesh out chroma key as an algorithm, in particular look at logical indexing as a way to pick out background pixels.

Start work on lab if time permits

Next Lecture