pl.beone.promena.transformer:barcode-detector-zxing-opencv

pl.beone.promena.sdk.maven.parent:promena-transformer-parent

License	Apache License, Version 2.0
Categories	OpenCV Business Logic Libraries Computer Vision ZXing Data Data Formats Image Processing ORM
GroupId	pl.beone.promena.transformer
ArtifactId	barcode-detector-zxing-opencv
Last Version	1.0.1
Release Date	Feb 6, 2020
Type	jar
Description	pl.beone.promena.sdk.maven.parent:promena-transformer-parent
Project URL	https://beone.pl
Project Organization	BeOne Sp. z o.o.
Source Code Management	https://github.com/BeOne-PL/promena-transformer-barcode-detector-zxing-opencv

Download barcode-detector-zxing-opencv

Filename	Size
barcode-detector-zxing-opencv-1.0.1.pom
barcode-detector-zxing-opencv-1.0.1.jar	113 KB
barcode-detector-zxing-opencv-1.0.1-sources.jar	17 KB
barcode-detector-zxing-opencv-1.0.1-javadoc.jar	2 KB

How to add to project

Apache Maven

<!-- https://jarcasting.com/artifacts/pl.beone.promena.transformer/barcode-detector-zxing-opencv/ -->
<dependency>
    <groupId>pl.beone.promena.transformer</groupId>
    <artifactId>barcode-detector-zxing-opencv</artifactId>
    <version>1.0.1</version>
</dependency>

Gradle Groovy

// https://jarcasting.com/artifacts/pl.beone.promena.transformer/barcode-detector-zxing-opencv/
implementation 'pl.beone.promena.transformer:barcode-detector-zxing-opencv:1.0.1'

Gradle Kotlin

// https://jarcasting.com/artifacts/pl.beone.promena.transformer/barcode-detector-zxing-opencv/
implementation ("pl.beone.promena.transformer:barcode-detector-zxing-opencv:1.0.1")

Apache Buildr

'pl.beone.promena.transformer:barcode-detector-zxing-opencv:jar:1.0.1'

Apache Ivy

<dependency org="pl.beone.promena.transformer" name="barcode-detector-zxing-opencv" rev="1.0.1">
  <artifact name="barcode-detector-zxing-opencv" type="jar" />
</dependency>

Groovy Grape

@Grapes(
@Grab(group='pl.beone.promena.transformer', module='barcode-detector-zxing-opencv', version='1.0.1')
)

Scala SBT

libraryDependencies += "pl.beone.promena.transformer" % "barcode-detector-zxing-opencv" % "1.0.1"

Leiningen

[pl.beone.promena.transformer/barcode-detector-zxing-opencv "1.0.1"]

Dependencies

compile (9)

Group / Artifact	Type	Version
org.jetbrains.kotlin : kotlin-stdlib-jdk8	jar	1.3.50
org.jetbrains.kotlin : kotlin-reflect	jar	1.3.50
pl.beone.promena.base : promena-transformer-util	jar	1.0.0
pl.beone.promena.transformer : barcode-detector-zxing-opencv-application-model	jar	1.0.1
io.github.microutils : kotlin-logging	jar	1.7.6
org.bytedeco : javacv-platform	jar	1.5.2
com.google.zxing : core	jar	3.4.0
com.google.zxing : javase	jar	3.4.0
org.apache.pdfbox : pdfbox	jar	2.0.16

test (5)

Group / Artifact	Type	Version
ch.qos.logback : logback-classic	jar	1.2.3
org.junit.jupiter : junit-jupiter-engine	jar	5.5.2
io.kotlintest : kotlintest-runner-junit5	jar	3.4.2
io.mockk : mockk	jar	1.9.3
pl.beone.lib : docker-extension-junit5	jar	1.0.0

Project Modules

There are no modules declared in this project.

Promena Transformer - barcode detector - ZXing & OpenCV

This transformer provides functionality to detect barcodes in a application/pdf document using ZXing 3.4.0 and OpenCV 4.1.2.

Visit Promena#Transformer to understand the repository structure.

The solution was developed thanks to the following resources:

• https://www.pyimagesearch.com/2014/11/24/detecting-barcodes-images-python-opencv/
• https://stackoverflow.com/a/50519348

Algorithm (for one page)

1. Convert to grayscale image with 300 DPI
2. Compute the Scharr gradient magnitude representation
3. Subtract the y-gradient from the x-gradient
4. Apply the binary threshold using values from linearThresholdValue and linearThresholdMaxValue parameters
5. Construct and apply the closing kernel using values from linearKernelSizeWidth and linearKernelSizeHeight parameters
6. Apply erosions and dilations using values from linearErosionsIterations and linearDilationsIterations parameters
7. Determine bounding boxes
8. Find the convex hull of each bounding box
9. If linearAdditionalVerticalTransformation is true, swap linearKernelSizeWidth parameter with linearKernelSizeHeight parameter and repeat steps 2-8
10. Cut out found contours to separate images
11. Calculate tilt of each image and rotation angles to rotate the image perpendicularly
12. After rotation apply additional rotations using value from linearRotationThresholdDegrees parameter
13. Try to detect a barcode on the image using formats from formats parameter
14. Repeat steps 2-13 for the set of parameters with matrix prefix
15. Filter found barcodes using value from regexFilter parameter

Transformation ZxingOpenCvBarcodeDetectorDsl, ZxingOpenCvBarcodeDetectorParametersDsl

The DataDescriptor has to contain at least one descriptor. If more than one descriptor is passed, the transformation will be performed on each of them separately.

Support ZxingOpenCvBarcodeDetectorSupport

Media type ZxingOpenCvBarcodeDetectorSupport.MediaTypeSupport

• application/pdf; UTF-8 ➡️ application/pdf; UTF-8

Parameters ZxingOpenCvBarcodeDetectorSupport.ParametersSupport

Take a look at Algorithm (for one page) to understand the meaning of given parameter.

• formats, List<String>, ["Codabar", "UPC-A", "UPC-E", "EAN-8", "EAN-13", "Code 39", "Code 128", "ITF", "RSS-14", "RSS Expanded", "QR Code", "PDF417", "Aztec Code", "Data Matrix", "MaxiCode"] optional
• regexFilter, String, optional
• linearRotationThresholdDegrees, Int, <1, 360>, optional
• linearAdditionalVerticalTransformation, Boolean, optional
• linearThresholdValue, Double, <1.0, 255.0>, optional
• linearThresholdMaxVal, Double, <1.0, 255.0>, optional
• linearKernelSizeWidth, Double, <1.0, 1000.0>, optional
• linearKernelSizeHeight, Double, <1.0, 1000.0>, optional
• linearErosionsIterations, Int, <1, 1000>, optional
• linearDilationsIterations, Int, <1, 1000>, optional
• matrixRotationThresholdDegrees, Int, <1, 360>, optional
• matrixAdditionalVerticalTransformation, Boolean, optional
• matrixThresholdValue, Double, <1.0, 255.0>, optional
• matrixThresholdMaxVal, Double, <1.0, 255.0>, optional
• matrixKernelSizeWidth, Double, <1.0, 1000.0>, optional
• matrixKernelSizeHeight, Double, <1.0, 1000.0>, optional
• matrixErosionsIterations, Int, <1, 1000>, optional
• matrixDilationsIterations, Int, <1, 1000>, optional

Metadata

Visit to find out how to build and process barcode metadata.

Dependency

<dependency>
    <groupId>pl.beone.promena.transformer</groupId>
    <artifactId>barcode-detector-zxing-opencv-configuration</artifactId>
    <version>1.0.1</version>
</dependency>

promena-docker-maven-plugin

<dependency>
    <groupId>pl.beone.promena.transformer</groupId>
    <artifactId>barcode-detector-zxing-opencv</artifactId>
    <version>1.0.1</version>
</dependency>

Properties

The default values are adjusted to recognise barcodes on scanned documents.

transformer.pl.beone.promena.transformer.barcodedetector.zxingopencv.ZxingOpenCvBarcodeDetectorTransformer.priority=1
transformer.pl.beone.promena.transformer.barcodedetector.zxingopencv.ZxingOpenCvBarcodeDetectorTransformer.actors=1

# RSS-14, RSS Expanded, UPC-E produces many false-positive detections
# UPC-A, EAN-8, EAN-13 also produces false-positive detections but less than previous ones
transformer.pl.beone.promena.transformer.barcodedetector.zxingopencv.default.parameters.formats=Codabar,Code 39,Code 128,ITF,QR Code,PDF417,Aztec Code,Data Matrix,MaxiCode
transformer.pl.beone.promena.transformer.barcodedetector.zxingopencv.default.parameters.regex-filter=

transformer.pl.beone.promena.transformer.barcodedetector.zxingopencv.default.parameters.linear-rotation-threshold-degrees=1
transformer.pl.beone.promena.transformer.barcodedetector.zxingopencv.default.parameters.linear-additional-vertical-transformation=true
transformer.pl.beone.promena.transformer.barcodedetector.zxingopencv.default.parameters.linear-threshold-value=150.0
transformer.pl.beone.promena.transformer.barcodedetector.zxingopencv.default.parameters.linear-threshold-max-val=255.0
transformer.pl.beone.promena.transformer.barcodedetector.zxingopencv.default.parameters.linear-kernel-size-width=25.0
transformer.pl.beone.promena.transformer.barcodedetector.zxingopencv.default.parameters.linear-kernel-size-height=1.0
transformer.pl.beone.promena.transformer.barcodedetector.zxingopencv.default.parameters.linear-erosions-iterations=20
transformer.pl.beone.promena.transformer.barcodedetector.zxingopencv.default.parameters.linear-dilations-iterations=20

transformer.pl.beone.promena.transformer.barcodedetector.zxingopencv.default.parameters.matrix-rotation-threshold-degrees=1
transformer.pl.beone.promena.transformer.barcodedetector.zxingopencv.default.parameters.matrix-additional-vertical-transformation=false
transformer.pl.beone.promena.transformer.barcodedetector.zxingopencv.default.parameters.matrix-threshold-value=100.0
transformer.pl.beone.promena.transformer.barcodedetector.zxingopencv.default.parameters.matrix-threshold-max-val=255.0
transformer.pl.beone.promena.transformer.barcodedetector.zxingopencv.default.parameters.matrix-kernel-size-width=20.0
transformer.pl.beone.promena.transformer.barcodedetector.zxingopencv.default.parameters.matrix-kernel-size-height=20.0
transformer.pl.beone.promena.transformer.barcodedetector.zxingopencv.default.parameters.matrix-erosions-iterations=20
transformer.pl.beone.promena.transformer.barcodedetector.zxingopencv.default.parameters.matrix-dilations-iterations=20
transformer.pl.beone.promena.transformer.barcodedetector.zxingopencv.default.parameters.timeout=

Evaluator

If you want to adjust parameters, run Main from barcode-detector-zxing-opencv-evaluator module (you have to install OpenCV 4.1.2 on your computer).

BeOne Sp. z o.o.

Versions

Version
1.0.1 Feb 6, 2020
1.0.0 Jan 23, 2020