Image Processing Toolbox

Working with uint8 and uint16 Data

Use imread to read graphics images into MATLAB as uint8 or uint16 arrays; use imshow to display these images; and use imwrite to save these images. Most of the functions in the Image Processing Toolbox accept uint8 and uint16 input. See the reference entries for more information about imread, imshow, uint8, and uint16.

MATLAB provides limited support for storing images as 8-bit or 16-bit unsigned integers. In addition to reading and writing uint8 and uint16 arrays, MATLAB supports the following operations:

• Displaying data values
• Indexing into arrays using standard MATLAB subscripting
• Reshaping, reordering, and concatenating arrays, using functions such as reshape, cat, and permute
• Saving to and loading from MAT-files
• The all and any functions
• Logical operators and indexing
• Relational operators
• The find function. Note that the returned array is of class double.

Mathematical Operations Support for uint8 and uint16

The following MATLAB mathematical operations support uint8 and uint16 data: conv2, convn, fft2, fftn, sum. In these cases, the output is always double.

If you attempt to perform an unsupported operation on one of these arrays, you will receive an error. For example,

BW3 = BW1 + BW2
??? Function '+' not defined for variables of class 'uint8'.

Converting The Image Type of Images

Converting The Storage Class of Images