Find Local Maxima

Find local maxima in matrices

• 

Libraries:
Computer Vision Toolbox / Statistics

Description

The Find Local Maxima block finds the local maxima in a given input matrix. The block finds the local maxima of values within a region specified by the neighborhood size. The block compares the values in the search region to a user-specified threshold and identifies a value that is greater than or equal to the threshold as the local maximum. After finding the local maximum, the block sets all the matrix values in the neighborhood, including the maximum value, to zero. This step ensures that subsequent searches do not include this maximum. The process repeats until the block either finds all valid maxima or the specified number of local maxima.

The threshold and the neighborhood size must be chosen such that the block detects only the true local maxima and eliminates false peaks.

The input to this block can also be a Hough matrix generated from the Hough Transform block. The Find Local Maxima block outputs the one-based [x y] coordinates of the detected local maxima. The data sent to all input ports must be the same data type.

Examples

Find Local Maxima in Image

Find local maxima in an image.

• Open Script

Ports

Input

I — Input

M-by-N matrix

Input, specified as the M-by-N matrix in which you want to detect the local maxima. If the input is a Hough matrix output from the Hough Transform block, select the Input is Hough matrix spanning full theta range parameter that renames this port as Hough.

Data Types: single | double | int8 | int16 | int32 | uint8 | uint16 | uint32 | fixed point

Th — Threshold for detecting local maxima

scalar

Threshold for the detecting local maxima, specified as a scalar.

Dependencies

To enable this port, set the Source of threshold value parameter to Input port.

Data Types: single | double | int8 | int16 | int32 | uint8 | uint16 | uint32 | fixed point

Output

Idx — Coordinates of local maxima

P-by-2 matrix

Coordinates of the local maxima, returned as a P-by-2 matrix of one-based [x y] coordinates. P represents the number of local maxima.

Data Types: single | double | uint8 | uint16 | uint32

Count — Number of local maxima

scalar

Number of local maxima, returned as a scalar.

Dependencies

To enable this port, clear the check box for the Output variable size signal parameter.

Data Types: single | double | uint8 | uint16 | uint32

Parameters

Maximum number of local maxima — Maximum number of local maxima to detect

2 (default) | positive integer

Specify the maximum number of local maxima to detect in the input matrix. The value must be a positive integer greater than 0.

Neighborhood size — Size of neighborhood region

[5 7] (default) | two-element vector

Specify the size of the neighborhood region as a two-element vector in the form [r c]. The vector specifies the window size r×c, within which the block determines the local maxima. The window size must be greater than or equal to 1×1.

Source of threshold value — Source to enter the threshold value

Specify via dialog (default) | Input port

Specify the source for entering the threshold value. When you select the Input port, the block enables the Th port. When you select the Specify via dialog, the block enables the Threshold parameter.

Threshold — Threshold for detecting local maxima

10 (default) | scalar

Specify a scalar value that represents the threshold for detecting the local maxima.

Dependencies

To enable this parameter, set the Source of threshold value parameter to Specify via dialog.

Input is Hough matrix spanning full theta range — Whether input is a Hough matrix

off (default) | on

Select the parameter, if the input to this block is the Hough matrix output from the Hough Transform block. The Find Local Maxima block assumes that the input is a Hough matrix that is antisymmetric about the rho axis, with values of theta from -pi/2 to pi/2 radians. As a result, if the local maximum is near the boundary and its neighborhood region is outside the Hough matrix, the block detects only one local maximum. The block ignores the corresponding antisymmetric maximum.

Index output data type — Data type for coordinates of local maximums

uint32 (default) | double | single | uint8 | uint16

Specify the data type for the coordinates of local maximums returned at the output port Idx.

Output variable size signal — Support input signals of varying length

on (default) | off

Specify if the model supports input signals of varying length.

Count output data type — Data type for number of local maximums

uint32 (default) | double | single | uint8 | uint16

Specify the data type for the number of local maximums returned at the output port Count.

Dependencies

To enable this parameter, clear the Output variable size signal parameter.

Block Characteristics

Data Types	double | fixed point | integer | single
Multidimensional Signals	no
Variable-Size Signals	yes

Extended Capabilities

C/C++ Code Generation
Generate C and C++ code using Simulink® Coder™.

Version History

Introduced before R2006a