hexagon.hvx.conv2d

R2026b

2-D convolution for HVX

Since R2026a

Syntax

conv2DOut = hexagon.hvx.conv2d(Input,Mask)

Description

conv2DOut = hexagon.hvx.conv2d(Input,Mask) performs a 2-D convolution of the input matrix Input with the structuring element Mask for Hexagon^® Vector eXtension.

Examples

2-D Convolution Using 3x3, 5x5, and 7x7 Kernels

Create a unit8 matrix. Specify the width of the input matrix as a multiple of 128.

H = 128; 
W = 128;                 
A = uint8(rand(H, W));

Create the mask with kernel sizes 3, 5, and 7.

mask3x3 = int8([ 1  2  1;0  0  0;-1 -2 -1 ]);
mask5x5 = int8([1  2  3  2  1;...
                2  4  5  4  2;...
                3  5  6  5  3;...
                2  4  5  4  2;...
                1  2  3  2  1]);
mask7x7 = int8([0  0  1  2  1  0  0;...
                0  1  2  3  2  1  0;...
                1  2  3  4  3  2  1;...
                2  3  4  5  4  3  2;...
                1  2  3  4  3  2  1;...
                0  1  2  3  2  1  0;...
                0  0  0  1  2  1  0]);

Perform convolution of the input matrix and the masks.

out3 = hexagon.hvx.conv2d(A, mask3x3);
out5 = hexagon.hvx.conv2d(A, mask5x5);
out7 = hexagon.hvx.conv2d(A, mask7x7);

Input Arguments

Input — 2-D input

matrix

2-D input, specified as a matrix whose width is a multiple of 128. The height of the matrix must be greater than or equal to the mask size.

Data Types: uint8

Mask — Structuring element

3-by-3 matrix (default) | 5-by-5 matrix | 7-by-7 matrix

Structuring element, specified as a 2-D square matrix of size 3, 5, or 7.

Data Types: int8

Output Arguments

conv2DOut — Convolution output

matrix

Convolution output of the input matrix and the mask, returned as a matrix.

Data Types: uint8

Version History

Introduced in R2026a

R2026b: QHVXIPCV.conv2d function is renamed to hexagon.hvx.conv2d

The QHVXIPCV.conv2d has been renamed to hexagon.hvx.conv2d.

If your code uses functions from the QHVXIPCV package, update your code to use the corresponding hexagon.hvx function instead.