/ 
Image-CCV-0.11
River stage zero No dependents
/ Image::CCV

NAME

Image::CCV - Crazy-cool Computer Vision bindings for Perl

SYNOPSIS

use Image::CCV qw(detect_faces);

my $scene = "image-with-faces.png";

my @coords = detect_faces( $scene );
print "@$_\n" for @coords;

ALPHA WARNING

This code is very, very rough. It leaks memory left and right and the API is very much in flux. But as I got easy results using this code already, I am releasing it as is and will improve it and the API as I go along.

See also http://libccv.org for the libccv website.

FUNCTIONS

default_sift_params(%options)

Sets up the parameter block for sift() and related routines. Valid keys for %options are:

  • noctaves - number of octaves

  • nlevels - number of levels

  • up2x - boolean, whether to upscale

  • edge_threshold - edge threshold

  • norm_threshold - norm threshold

  • peak_threshold - peak threshold

get_sift_descriptor( $image, $parameters );

my $desc = get_sift_descriptor('image.png');
print for @{ $desc->{keypoints} };

Not yet implemented

sift( $object, $scene, $params )

my @common_features = sift( 'object.png', 'sample.png' );

Returns a list of 4-element arrayrefs. The elements are:

object-x
object-y
scene-x
scene-y

The parameters get decoded by get_default_params.

detect_faces( $png_file )

my @faces = detect_faces('sample.png');

Returns a list of 5-element arrayrefs. The elements are:

x
y
width
height
confidence

LIMITATIONS

Due to the early development stages, there are several limitations.

Limited data transfer

Currently, the only mechanism to pass in image data to ccv is by loading grayscale PNG or JPEG images from disk. The plan is to also be able to pass in image data as scalars or Imager objects.

Limited result storage

Currently, there is no implemented way to store the results of applying the SIFT algorithm to an image. This makes searching several images for the same object slow and inconvenient.

Limited memory management

Memory currently is only allocated. Rarely is memory deallocated.

REPOSITORY

The public repository of this module is https://github.com/Corion/image-ccv.

The upstream repository of libccv is at

SUPPORT

The public support forum of this module is https://perlmonks.org/.

The support for libccv can be found at http://libccv.org.

TALKS

I've given one lightning talk about this module at Perl conferences:

German Perl Workshop, German

BUG TRACKER

Please report bugs in this module via the RT CPAN bug queue at https://rt.cpan.org/Public/Dist/Display.html?Name=Image-CCV or via mail to image-ccv-Bugs@rt.cpan.org.

INSTALL

Compilation requires -dev header libraries, so make sure you have (at the time of writing, on *nix) libjpeg8-dev and libpng12-dev installed.

AUTHOR

Max Maischein corion@cpan.org

COPYRIGHT (c)

Copyright 2012-2017 by Max Maischein corion@cpan.org.

LICENSE

This module is released under the same terms as Perl itself. The CCV library distributed with it comes with its own license(s). Please study these before redistributing.