PHP中的人脸检测 [英] Face detection in PHP
本文介绍了PHP中的人脸检测的处理方法，对大家解决问题具有一定的参考价值，需要的朋友们下面随着小编来一起学习吧！
问题描述
有人知道在PHP中进行人脸检测的好方法吗？ 我遇到过这里的一些代码声称这样做，但我似乎无法让它正常工作。我想做这个工作（即使它会很慢），你可以给我的任何帮助都会非常感激。
Does anybody know of a good way to do face detection in PHP? I came across some code here that claims to do this, but I can't seem to get it to work properly. I'd like to make this work (even though it will be slow) and any help you can give me would be really appreciated.
这是链接中的代码：
<?php
// as published by the Free Software Foundation; either version 2
// of the License, or (at your option) any later version.
//
// This program is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU General Public License for more details.
//
// You should have received a copy of the GNU General Public License
// along with this program; if not, write to the Free Software
// Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
//
// @Author Karthik Tharavaad
// karthik_tharavaad@yahoo.com
// @Contributor Maurice Svay
// maurice@svay.Com
class Face_Detector {
protected $detection_data;
protected $canvas;
protected $face;
private $reduced_canvas;
public function __construct($detection_file = 'detection.dat') {
if (is_file($detection_file)) {
$this->detection_data = unserialize(file_get_contents($detection_file));
} else {
throw new Exception("Couldn't load detection data");
}
//$this->detection_data = json_decode(file_get_contents('data.js'));
}
public function face_detect($file) {
if (!is_file($file)) {
throw new Exception("Can not load $file");
}
$this->canvas = imagecreatefromjpeg($file);
$im_width = imagesx($this->canvas);
$im_height = imagesy($this->canvas);
//Resample before detection?
$ratio = 0;
$diff_width = 320 - $im_width;
$diff_height = 240 - $im_height;
if ($diff_width > $diff_height) {
$ratio = $im_width / 320;
} else {
$ratio = $im_height / 240;
}
if ($ratio != 0) {
$this->reduced_canvas = imagecreatetruecolor($im_width / $ratio, $im_height / $ratio);
imagecopyresampled($this->reduced_canvas, $this->canvas, 0, 0, 0, 0, $im_width / $ratio, $im_height / $ratio, $im_width, $im_height);
$stats = $this->get_img_stats($this->reduced_canvas);
$this->face = $this->do_detect_greedy_big_to_small($stats['ii'], $stats['ii2'], $stats['width'], $stats['height']);
$this->face['x'] *= $ratio;
$this->face['y'] *= $ratio;
$this->face['w'] *= $ratio;
} else {
$stats = $this->get_img_stats($this->canvas);
$this->face = $this->do_detect_greedy_big_to_small($stats['ii'], $stats['ii2'], $stats['width'], $stats['height']);
}
return ($this->face['w'] > 0);
}
public function toJpeg() {
$color = imagecolorallocate($this->canvas, 255, 0, 0); //red
imagerectangle($this->canvas, $this->face['x'], $this->face['y'], $this->face['x']+$this->face['w'], $this->face['y']+ $this->face['w'], $color);
header('Content-type: image/jpeg');
imagejpeg($this->canvas);
}
public function toJson() {
return "{'x':" . $this->face['x'] . ", 'y':" . $this->face['y'] . ", 'w':" . $this->face['w'] . "}";
}
public function getFace() {
return $this->face;
}
protected function get_img_stats($canvas){
$image_width = imagesx($canvas);
$image_height = imagesy($canvas);
$iis = $this->compute_ii($canvas, $image_width, $image_height);
return array(
'width' => $image_width,
'height' => $image_height,
'ii' => $iis['ii'],
'ii2' => $iis['ii2']
);
}
protected function compute_ii($canvas, $image_width, $image_height ){
$ii_w = $image_width+1;
$ii_h = $image_height+1;
$ii = array();
$ii2 = array();
for($i=0; $i<$ii_w; $i++ ){
$ii[$i] = 0;
$ii2[$i] = 0;
}
for($i=1; $i<$ii_w; $i++ ){
$ii[$i*$ii_w] = 0;
$ii2[$i*$ii_w] = 0;
$rowsum = 0;
$rowsum2 = 0;
for($j=1; $j<$ii_h; $j++ ){
$rgb = ImageColorAt($canvas, $j, $i);
$red = ($rgb >> 16) & 0xFF;
$green = ($rgb >> 8) & 0xFF;
$blue = $rgb & 0xFF;
$grey = ( 0.2989*$red + 0.587*$green + 0.114*$blue )>>0; // this is what matlab uses
$rowsum += $grey;
$rowsum2 += $grey*$grey;
$ii_above = ($i-1)*$ii_w + $j;
$ii_this = $i*$ii_w + $j;
$ii[$ii_this] = $ii[$ii_above] + $rowsum;
$ii2[$ii_this] = $ii2[$ii_above] + $rowsum2;
}
}
return array('ii'=>$ii, 'ii2' => $ii2);
}
protected function do_detect_greedy_big_to_small( $ii, $ii2, $width, $height ){
$s_w = $width/20.0;
$s_h = $height/20.0;
$start_scale = $s_h < $s_w ? $s_h : $s_w;
$scale_update = 1 / 1.2;
for($scale = $start_scale; $scale > 1; $scale *= $scale_update ){
$w = (20*$scale) >> 0;
$endx = $width - $w - 1;
$endy = $height - $w - 1;
$step = max( $scale, 2 ) >> 0;
$inv_area = 1 / ($w*$w);
for($y = 0; $y < $endy ; $y += $step ){
for($x = 0; $x < $endx ; $x += $step ){
$passed = $this->detect_on_sub_image( $x, $y, $scale, $ii, $ii2, $w, $width+1, $inv_area);
if( $passed ) {
return array('x'=>$x, 'y'=>$y, 'w'=>$w);
}
} // end x
} // end y
} // end scale
return null;
}
protected function detect_on_sub_image( $x, $y, $scale, $ii, $ii2, $w, $iiw, $inv_area){
$mean = ( $ii[($y+$w)*$iiw + $x + $w] + $ii[$y*$iiw+$x] - $ii[($y+$w)*$iiw+$x] - $ii[$y*$iiw+$x+$w] )*$inv_area;
$vnorm = ( $ii2[($y+$w)*$iiw + $x + $w] + $ii2[$y*$iiw+$x] - $ii2[($y+$w)*$iiw+$x] - $ii2[$y*$iiw+$x+$w] )*$inv_area - ($mean*$mean);
$vnorm = $vnorm > 1 ? sqrt($vnorm) : 1;
$passed = true;
for($i_stage = 0; $i_stage < count($this->detection_data); $i_stage++ ){
$stage = $this->detection_data[$i_stage];
$trees = $stage[0];
$stage_thresh = $stage[1];
$stage_sum = 0;
for($i_tree = 0; $i_tree < count($trees); $i_tree++ ){
$tree = $trees[$i_tree];
$current_node = $tree[0];
$tree_sum = 0;
while( $current_node != null ){
$vals = $current_node[0];
$node_thresh = $vals[0];
$leftval = $vals[1];
$rightval = $vals[2];
$leftidx = $vals[3];
$rightidx = $vals[4];
$rects = $current_node[1];
$rect_sum = 0;
for( $i_rect = 0; $i_rect < count($rects); $i_rect++ ){
$s = $scale;
$rect = $rects[$i_rect];
$rx = ($rect[0]*$s+$x)>>0;
$ry = ($rect[1]*$s+$y)>>0;
$rw = ($rect[2]*$s)>>0;
$rh = ($rect[3]*$s)>>0;
$wt = $rect[4];
$r_sum = ( $ii[($ry+$rh)*$iiw + $rx + $rw] + $ii[$ry*$iiw+$rx] - $ii[($ry+$rh)*$iiw+$rx] - $ii[$ry*$iiw+$rx+$rw] )*$wt;
$rect_sum += $r_sum;
}
$rect_sum *= $inv_area;
$current_node = null;
if( $rect_sum >= $node_thresh*$vnorm ){
if( $rightidx == -1 )
$tree_sum = $rightval;
else
$current_node = $tree[$rightidx];
} else {
if( $leftidx == -1 )
$tree_sum = $leftval;
else
$current_node = $tree[$leftidx];
}
}
$stage_sum += $tree_sum;
}
if( $stage_sum < $stage_thresh ){
return false;
}
}
return true;
}
}
用法：
$detector = new Face_Detector('detection.dat');
$detector->face_detect('maurice_svay_150.jpg');
$detector->toJpeg();
我遇到的问题似乎也出现在该页面的评论中。 imagecolorat（）[function.imagecolorat]：320,1超出范围。所以，我在文件的顶部添加了一个error_reporting（0）（实际上不是解决方案），它似乎有时会工作，而其他时候它什么都不做。
The problem I am running into, seems to be coming up in the comments on that page as well. "imagecolorat() [function.imagecolorat]: 320,1 is out of bounds." So, I added a error_reporting(0) to the top of the file (not really the solution), and it seems to work sometimes while other times it just doesn't do anything.
有什么想法？
推荐答案
您需要关闭错误报告
<?php
ini_set( 'display_errors', 1 );
error_reporting( E_ALL ^ E_NOTICE );
require_once('face_detector.php');
$detector = new Face_Detector('detection.dat');
$detector->face_detect('img/8.jpg');
$detector->toJpeg();
?>
这篇关于PHP中的人脸检测的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！
查看全文
