$gridSampler
$gridSampler
Implementations of this class can, given locations of finder patterns for a QR code in an image, sample the right points in the image to reconstruct the QR code, accounting for perspective distortion. It is abstracted since it is relatively expensive and should be allowed to take advantage of platform-specific optimized implementations, like Sun's Java Advanced Imaging library, but which may not be available in other environments such as J2ME, and vice versa.
setGridSampler(mixed $newGridSampler) : mixed
Sets the implementation of GridSampler used by the library. One global instance is stored, which may sound problematic. But, the implementation provided ought to be appropriate for the entire platform, and all uses of this library in the whole lifetime of the JVM. For instance, an Android activity can swap in an implementation that takes advantage of native platform libraries.
mixed | $newGridSampler |
sampleGrid(mixed $image, mixed $dimensionX, mixed $dimensionY, mixed $p1ToX, mixed $p1ToY, mixed $p2ToX, mixed $p2ToY, mixed $p3ToX, mixed $p3ToY, mixed $p4ToX, mixed $p4ToY, mixed $p1FromX, mixed $p1FromY, mixed $p2FromX, mixed $p2FromY, mixed $p3FromX, mixed $p3FromY, mixed $p4FromX, mixed $p4FromY) : mixed
Samples an image for a rectangular matrix of bits of the given dimension. The sampling transformation is determined by the coordinates of 4 points, in the original and transformed image space.
mixed | $image | |
mixed | $dimensionX | |
mixed | $dimensionY | |
mixed | $p1ToX | |
mixed | $p1ToY | |
mixed | $p2ToX | |
mixed | $p2ToY | |
mixed | $p3ToX | |
mixed | $p3ToY | |
mixed | $p4ToX | |
mixed | $p4ToY | |
mixed | $p1FromX | |
mixed | $p1FromY | |
mixed | $p2FromX | |
mixed | $p2FromY | |
mixed | $p3FromX | |
mixed | $p3FromY | |
mixed | $p4FromX | |
mixed | $p4FromY |
checkAndNudgePoints(mixed $image, mixed $points) : mixed
<p>Checks a set of points that have been transformed to sample points on an image against the image's dimensions to see if the point are even within the image.</p>
This method will actually "nudge" the endpoints back onto the image if they are found to be barely (less than 1 pixel) off the image. This accounts for imperfect detection of finder patterns in an image where the QR Code runs all the way to the image border.
For efficiency, the method will check points from either end of the line until one is found to be within the image. Because the set of points are assumed to be linear, this is valid.
mixed | $image | |
mixed | $points |
if an endpoint is lies outside the image boundaries
<?php
/*
* Copyright 2007 ZXing authors
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
namespace Zxing\Common;
use Zxing\NotFoundException;
/**
* @author Sean Owen
*/
final class DefaultGridSampler extends GridSampler
{
//@Override
public function sampleGrid(
$image,
$dimensionX,
$dimensionY,
$p1ToX, $p1ToY,
$p2ToX, $p2ToY,
$p3ToX, $p3ToY,
$p4ToX, $p4ToY,
$p1FromX, $p1FromY,
$p2FromX, $p2FromY,
$p3FromX, $p3FromY,
$p4FromX, $p4FromY
) {
$transform = PerspectiveTransform::quadrilateralToQuadrilateral(
$p1ToX, $p1ToY, $p2ToX, $p2ToY, $p3ToX, $p3ToY, $p4ToX, $p4ToY,
$p1FromX, $p1FromY, $p2FromX, $p2FromY, $p3FromX, $p3FromY, $p4FromX, $p4FromY);
return $this->sampleGrid_($image, $dimensionX, $dimensionY, $transform);
}
//@Override
public function sampleGrid_(
$image,
$dimensionX,
$dimensionY,
$transform
) {
if ($dimensionX <= 0 || $dimensionY <= 0) {
throw NotFoundException::getNotFoundInstance();
}
$bits = new BitMatrix($dimensionX, $dimensionY);
$points = fill_array(0, 2 * $dimensionX, 0.0);
for ($y = 0; $y < $dimensionY; $y++) {
$max = count($points);
$iValue = (float)$y + 0.5;
for ($x = 0; $x < $max; $x += 2) {
$points[$x] = (float)($x / 2) + 0.5;
$points[$x + 1] = $iValue;
}
$transform->transformPoints($points);
// Quick check to see if points transformed to something inside the image;
// sufficient to check the endpoints
$this->checkAndNudgePoints($image, $points);
try {
for ($x = 0; $x < $max; $x += 2) {
if ($image->get((int)$points[$x], (int)$points[$x + 1])) {
// Black(-ish) pixel
$bits->set($x / 2, $y);
}
}
} catch (\Exception $aioobe) {//ArrayIndexOutOfBoundsException
// This feels wrong, but, sometimes if the finder patterns are misidentified, the resulting
// transform gets "twisted" such that it maps a straight line of points to a set of points
// whose endpoints are in bounds, but others are not. There is probably some mathematical
// way to detect this about the transformation that I don't know yet.
// This results in an ugly runtime exception despite our clever checks above -- can't have
// that. We could check each point's coordinates but that feels duplicative. We settle for
// catching and wrapping ArrayIndexOutOfBoundsException.
throw NotFoundException::getNotFoundInstance();
}
}
return $bits;
}
}