Added oneshot apps

2024-01-10 23:33:58 +01:00
parent 946eea2347
commit 6eda25fce5
24 changed files with 192 additions and 0 deletions
@@ -0,0 +1,13 @@
+# Default ignored files
+/shelf/
+/workspace.xml
+# Rider ignored files
+/contentModel.xml
+/projectSettingsUpdater.xml
+/modules.xml
+/.idea.Implementation.iml
+# Editor-based HTTP Client requests
+/httpRequests/
+# Datasource local storage ignored files
+/dataSources/
+/dataSources.local.xml
@@ -0,0 +1,4 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<project version="4">
+  <component name="Encoding" addBOMForNewFiles="with BOM under Windows, with no BOM otherwise" />
+</project>
@@ -0,0 +1,15 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<project version="4">
+  <component name="GitToolBoxProjectSettings">
+    <option name="commitMessageIssueKeyValidationOverride">
+      <BoolValueOverride>
+        <option name="enabled" value="true" />
+      </BoolValueOverride>
+    </option>
+    <option name="commitMessageValidationEnabledOverride">
+      <BoolValueOverride>
+        <option name="enabled" value="true" />
+      </BoolValueOverride>
+    </option>
+  </component>
+</project>
@@ -0,0 +1,8 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<project version="4">
+  <component name="UserContentModel">
+    <attachedFolders />
+    <explicitIncludes />
+    <explicitExcludes />
+  </component>
+</project>
@@ -0,0 +1,7 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<project version="4">
+  <component name="DiscordProjectSettings">
+    <option name="show" value="ASK" />
+    <option name="description" value="" />
+  </component>
+</project>
@@ -0,0 +1,6 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<project version="4">
+  <component name="VcsDirectoryMappings">
+    <mapping directory="$PROJECT_DIR$/.." vcs="Git" />
+  </component>
+</project>
@@ -45,6 +45,10 @@ Project("{9A19103F-16F7-4668-BE54-9A1E7A4F7556}") = "ReportGeneration.Generators
 EndProject
 Project("{9A19103F-16F7-4668-BE54-9A1E7A4F7556}") = "ReportGeneration.Abstract", "..\ReportGeneration\ReportGeneration.Abstract\ReportGeneration.Abstract.csproj", "{80BF15A5-78EB-4B87-8C1B-7F90D6D8BC74}"
 EndProject
+Project("{FAE04EC0-301F-11D3-BF4B-00C04F79EFBC}") = "Ocr.OneShot.Resampling", "Ocr.OneShot.Resampling\Ocr.OneShot.Resampling.csproj", "{4149E8A8-BD3C-4581-BDDD-FA70A58FBE37}"
+EndProject
+Project("{FAE04EC0-301F-11D3-BF4B-00C04F79EFBC}") = "Ocr.OneShot.Postprocessing", "Ocr.OneShot.Postprocessing\Ocr.OneShot.Postprocessing.csproj", "{40EA7F8E-BF46-410F-BA2B-CB05B80DCDBB}"
+EndProject
 Global
 	GlobalSection(SolutionConfigurationPlatforms) = preSolution
 		Debug|Any CPU = Debug|Any CPU
@@ -115,6 +119,14 @@ Global
 		{80BF15A5-78EB-4B87-8C1B-7F90D6D8BC74}.Debug|Any CPU.Build.0 = Debug|Any CPU
 		{80BF15A5-78EB-4B87-8C1B-7F90D6D8BC74}.Release|Any CPU.ActiveCfg = Release|Any CPU
 		{80BF15A5-78EB-4B87-8C1B-7F90D6D8BC74}.Release|Any CPU.Build.0 = Release|Any CPU
+		{4149E8A8-BD3C-4581-BDDD-FA70A58FBE37}.Debug|Any CPU.ActiveCfg = Debug|Any CPU
+		{4149E8A8-BD3C-4581-BDDD-FA70A58FBE37}.Debug|Any CPU.Build.0 = Debug|Any CPU
+		{4149E8A8-BD3C-4581-BDDD-FA70A58FBE37}.Release|Any CPU.ActiveCfg = Release|Any CPU
+		{4149E8A8-BD3C-4581-BDDD-FA70A58FBE37}.Release|Any CPU.Build.0 = Release|Any CPU
+		{40EA7F8E-BF46-410F-BA2B-CB05B80DCDBB}.Debug|Any CPU.ActiveCfg = Debug|Any CPU
+		{40EA7F8E-BF46-410F-BA2B-CB05B80DCDBB}.Debug|Any CPU.Build.0 = Debug|Any CPU
+		{40EA7F8E-BF46-410F-BA2B-CB05B80DCDBB}.Release|Any CPU.ActiveCfg = Release|Any CPU
+		{40EA7F8E-BF46-410F-BA2B-CB05B80DCDBB}.Release|Any CPU.Build.0 = Release|Any CPU
 	EndGlobalSection
 	GlobalSection(SolutionProperties) = preSolution
 		HideSolutionNode = FALSE
@@ -0,0 +1,15 @@
+<Project Sdk="Microsoft.NET.Sdk">
+
+    <PropertyGroup>
+        <OutputType>Exe</OutputType>
+        <TargetFramework>net7.0</TargetFramework>
+        <ImplicitUsings>enable</ImplicitUsings>
+        <Nullable>enable</Nullable>
+    </PropertyGroup>
+
+    <ItemGroup>
+      <ProjectReference Include="..\..\Ocr\Ocr.Tesseract.Screenshots\Ocr.Tesseract.Screenshots.csproj" />
+      <ProjectReference Include="..\..\Ocr\Ocr.Tesseract\Ocr.Tesseract.csproj" />
+    </ItemGroup>
+
+</Project>
@@ -0,0 +1,51 @@
+// See https://aka.ms/new-console-template for more information
+
+using System.Text.RegularExpressions;
+using ImageMagick;
+using Ocr.Tesseract;
+using Ocr.Tesseract.Models;
+using Ocr.Tesseract.Screenshots;
+using Ocr.Tesseract.Screenshots.Configuration;
+using Ocr.Tesseract.Screenshots.Threshold;
+using Process.Abstract.Configuration;
+using Process.Interface;
+
+
+Console.WriteLine("Hello, World!");
+
+IProcessorChain<MagickImage, ScanResult> MakeProcessor()
+{
+    var wordRegex = new Regex(
+        @"[\w'\-]{2,}",
+        RegexOptions.Compiled | RegexOptions.Multiline | RegexOptions.IgnoreCase
+    );
+    
+    var tesseractConfig = new TesseractScreenshotConfiguration
+    {
+        DataPath = "tessdata",
+        Languages = new[] { "eng", "deu" }
+    };
+    
+    var preprocessing = new ProcessorChainConfiguration<MagickImage, MagickImage>()
+        .Use(new CloneImageProcessor())
+        .Use(new ResizeProcessor(FilterType.Lanczos2Sharp, PixelInterpolateMethod.Mesh))
+        .Use(new NormalizeProcessor())
+        .Use(new ThresholdAdaptiveProcessor(15, 15))
+        .Use(new AddBorderProcessor(10))
+        .Use(new BinarizeProcessor())
+        .Complete(new NegateCloneProcessor());
+
+    var postprocessing = new ProcessorChainConfiguration<ScanResult, ScanResult>()
+        .Use(new ConfidenceFilter(50))
+        // todo insert processing events and write to json files
+        .Use(new ToLowerProcessor())
+        .Use(new DuplicateFilter())
+        .Complete(new RegexFilter(wordRegex));
+
+    var scan = new TesseractProcessor(tesseractConfig);
+
+    return new ProcessorChainConfiguration<MagickImage, ScanResult>()
+        .Use(preprocessing)
+        .Use(scan)
+        .Complete(postprocessing);
+}
@@ -0,0 +1,18 @@
+<Project Sdk="Microsoft.NET.Sdk">
+
+    <PropertyGroup>
+        <OutputType>Exe</OutputType>
+        <TargetFramework>net7.0</TargetFramework>
+        <ImplicitUsings>enable</ImplicitUsings>
+        <Nullable>enable</Nullable>
+    </PropertyGroup>
+
+    <ItemGroup>
+      <PackageReference Include="Magick.NET-Q16-AnyCPU" Version="13.2.0" />
+    </ItemGroup>
+
+    <ItemGroup>
+      <ProjectReference Include="..\..\Ocr\Ocr.Tesseract\Ocr.Tesseract.csproj" />
+    </ItemGroup>
+
+</Project>
@@ -0,0 +1,29 @@
+using ImageMagick;
+using Ocr.Tesseract.Extensions;
+
+const float factor = 4f;
+const string downscaled = "downscaled.png";
+
+using (var image = new MagickImage(args.Single()))
+{
+    image.ResizeImage(0.8f, FilterType.Catrom, PixelInterpolateMethod.Undefined);
+    image.Write(downscaled);
+}
+
+using (var image = new MagickImage(downscaled))
+{
+    image.ResizeImage(factor, FilterType.Lanczos, PixelInterpolateMethod.Catrom);
+    image.Write("Lanczos.png");
+}
+
+using (var image = new MagickImage(downscaled))
+{
+    image.ResizeImage(factor, FilterType.Point, PixelInterpolateMethod.Integer);
+    image.Write("Nearest.png");
+}
+
+using (var image = new MagickImage(downscaled))
+{
+    image.ResizeImage(factor, FilterType.Hermite, PixelInterpolateMethod.Mesh);
+    image.Write("Hermite.png");
+}
@@ -0,0 +1,14 @@
+{
+  "profiles": {
+    "Test single img": {
+      "commandName": "Project",
+      "commandLineArgs": "source.png",
+      "workingDirectory": "D:\\git\\ba_ocr\\ocr\\Implementation\\testdata"
+    },
+    "Refresh thesis results": {
+      "commandName": "Project",
+      "commandLineArgs": "source.png",
+      "workingDirectory": "C:\\Users\\Simon\\Documents\\Userdata\\FH\\SEM5\\BA\\bsc\\include\\resampling"
+    }
+  }
+}