Basic files for Visualization assignment

modules/cppcourse/cppcourse.cmake

+# modules/cppcourse/cppcourse.cmake
+# CMake file for the Module for the Cpp Course in WS 17
+
+# Set module status (valid values are STABLE, TESTING and EXPERIMENTAL)
+SET(ThisModStatus EXPERIMENTAL)
+# Set whether this module has external dependencies that are not shipped with CAMPVis.
+SET(ThisModExternalDependencies TRUE)
+
+# The files and assignments need only to be parsed if the module is enabled
+IF(ModuleEnabled)
+	# Source files:
+	FILE(GLOB ThisModSources RELATIVE ${ModulesDir}
+		modules/cppcourse/pipelines/*.cpp
+		modules/cppcourse/processors/*.cpp
+		modules/cppcourse/datastructures/*.cpp
+		modules/cppcourse/properties/*.cpp
+		modules/cppcourse/*.cpp
+	)
+
+	# Header files (including GLSL files so that they'll appear in VS projects)
+	FILE(GLOB ThisModHeaders RELATIVE ${ModulesDir}
+		modules/cppcourse/glsl/*.frag
+		modules/cppcourse/glsl/*.geom
+		modules/cppcourse/glsl/*.vert
+		modules/cppcourse/pipelines/*.h
+		modules/cppcourse/processors/*.h
+		modules/cppcourse/datastructures/*.h
+		modules/cppcourse/properties/*.h
+	)
+
+	# Define the GLSL shader path, so that all needed shaders will be deployed to target directory
+	SET(ThisModShaderDirectories "modules/cppcourse/glsl")
+
+	# Define dependency modules (the pipelines in the vis modules use processors from io)
+	#SET(ThisModDependencies io)
+	
+ENDIF(ModuleEnabled)

modules/cppcourse/cppcourse.cpp

+#include "core/pipeline/pipelinefactory.h"
+#include "core/pipeline/processorfactory.h"
+
+#include "modules/cppcourse/pipelines/cppcourse.h"
+#include "modules/cppcourse/processors/verysimpleraycaster.h"
+
+namespace campvis {
+
+    // explicitly instantiate templates to register the pipelines to make them available in the prototyping GUI
+    //template class PipelineRegistrar<CPPCourse>;
+
+    template class SmartProcessorRegistrar<VerySimpleRaycaster>;
+}

modules/cppcourse/glsl/verysimpleraycaster.frag

+// ================================================================================================
+// 
+// This file is part of the CAMPVis Software Framework.
+// 
+// If not explicitly stated otherwise: Copyright (C) 2012-2015, all rights reserved,
+//      Christian Schulte zu Berge <christian.szb@in.tum.de>
+//      Chair for Computer Aided Medical Procedures
+//      Technische Universitaet Muenchen
+//      Boltzmannstr. 3, 85748 Garching b. Muenchen, Germany
+// 
+// For a full list of authors and contributors, please refer to the file "AUTHORS.txt".
+// 
+// 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.
+// 
+// ================================================================================================
+
+layout(location = 0) out vec4 FragData0;     ///< outgoing fragment color
+
+
+#include "tools/gradient.frag"
+#include "tools/raycasting.frag"
+#include "tools/shading.frag"
+#include "tools/texture2d.frag"
+#include "tools/texture3d.frag"
+#include "tools/transferfunction.frag"
+
+uniform vec2 _viewportSizeRCP;
+uniform float _jitterStepSizeMultiplier;
+
+
+// ray entry points
+uniform sampler2D _entryPoints;
+uniform sampler2D _entryPointsDepth;
+uniform TextureParameters2D _entryParams;
+
+// ray exit points
+uniform sampler2D _exitPoints;
+uniform sampler2D _exitPointsDepth;
+uniform TextureParameters2D _exitParams;
+
+// DVR volume
+uniform sampler3D _volume;
+uniform TextureParameters3D _volumeTextureParams;
+
+// Transfer function
+uniform sampler1D _transferFunction;
+uniform TFParameters1D _transferFunctionParams;
+
+
+uniform LightSource _lightSource;
+uniform vec3 _cameraPosition;
+
+uniform float _samplingStepSize;
+
+const float SAMPLING_BASE_INTERVAL_RCP = 200.0;
+
+// The compositing mode that gets passed through from the p_compositingMode property
+uniform int _compositingMode;
+
+/**
+ * Performs the raycasting and returns the final fragment color.
+ */
+vec4 performRaycasting(in vec3 entryPoint, in vec3 exitPoint, in vec2 texCoords) {
+    vec4 result = vec4(0.0);
+    float firstHitT = -1.0;
+
+    // calculate ray start and endpoint
+    vec3 direction = exitPoint.rgb - entryPoint.rgb;
+    float t = 0.0;
+    float tend = length(direction);
+    direction = normalize(direction);
+
+    // This applies a small random offset to avoid banding artifacts
+    // Uncomment to see the effect ;)
+    jitterEntryPoint(entryPoint, direction, _samplingStepSize * _jitterStepSizeMultiplier);
+
+    while (t < tend) {
+        // compute sample position - next step
+        vec3 samplePosition = entryPoint.rgb + t * direction;
+
+        // get the intensity of our sample from the volume
+        float intensity = texture(_volume, samplePosition).r;
+
+        // use the transfer function to transform from 
+        vec4 color = lookupTF(_transferFunction, _transferFunctionParams, intensity);
+
+        if (_compositingMode == 0) {
+            // accomodate for variable sampling rates
+            color.a = 1.0 - pow(1.0 - color.a, _samplingStepSize * SAMPLING_BASE_INTERVAL_RCP);
+
+            // perform compositing
+            // Alpha-blend the new sample with the color accumulated so far
+            blendUnder(result, color);
+        }
+        else if (_compositingMode == 1) {
+            //TODO
+            result = vec4(entryPoint, 1.0);
+        }
+        // else if...
+
+        // save first hit ray parameter for depth value calculation
+        if (firstHitT < 0.0 && result.a > 0.0) {
+            firstHitT = t;
+        }
+
+        // advance to the next evaluation point along the ray
+        t += _samplingStepSize;
+    }
+
+    // calculate depth value from ray parameter
+    gl_FragDepth = 1.0;
+    if (firstHitT >= 0.0) {
+        float depthEntry = texture(_entryPointsDepth, texCoords).z;
+        float depthExit = texture(_exitPointsDepth, texCoords).z;
+        gl_FragDepth = calculateDepthValue(firstHitT/tend, depthEntry, depthExit);
+    }
+    return result;
+}
+
+/***
+ * The main method.
+ ***/
+void main() {
+    vec2 p = gl_FragCoord.xy * _viewportSizeRCP;
+    vec3 frontPos = texture(_entryPoints, p).rgb;
+    vec3 backPos = texture(_exitPoints, p).rgb;
+
+    //determine whether the ray has to be casted
+    if (frontPos == backPos) {
+        //background needs no raycasting
+        discard;
+    } else {
+        //fragCoords are lying inside the boundingbox
+        FragData0 = performRaycasting(frontPos, backPos, p);
+    }
+}

modules/cppcourse/pipelines/cppcourse.cpp

+#include "cppcourse.h"
+
+// TODO

modules/cppcourse/pipelines/cppcourse.h

+#pragma once
+
+// TODO

modules/cppcourse/processors/EDFhandler.cpp

+#include "EDFhandler.h"
+
+#include <fstream>
+#include <map>
+#include <exception>
+
+#include "IOutilities.h"
+
+
+Volume EDFHandler::read(const std::string& filename)
+{
+    // open the file
+    std::ifstream file(filename, std::ios::binary);
+    if (!file.good())
+        throw std::invalid_argument("EDFHandler::read: error opening " + filename + " for reading");
+
+    // read a single character and make sure that a header is opened
+    if (file.eof() || file.get() != '{')
+        throw std::invalid_argument("EDFHandler::read: failed reading header opening marker");
+
+    std::map<std::string, std::string> attributes;
+
+    // --- read header data
+    while (!file.eof()) {
+        // skip whitespace
+        while (!file.eof()) {
+            const int chr = file.peek();
+            if (chr != '\r' && chr != '\n' && chr != ' ' && chr != '\t')
+                break;
+            file.ignore(1);
+        }
+
+        // abort if the header is closed
+        if (file.eof() || file.peek() == '}')
+            break;
+
+        // extract the attribute assignment
+        bool quotesSingle = false, quotesDouble = false;
+        std::string assignment = "";
+        while (!file.eof()) {
+            const int chr = file.get();
+
+            // abort on end-of-assignment
+            if (chr == ';' && !(quotesSingle || quotesDouble))
+                break;
+
+            // check for quote characters
+            if (chr == '\'') quotesSingle = !quotesSingle;
+            if (chr == '\"') quotesDouble = !quotesDouble;
+
+            assignment += chr;
+        }
+
+        // split the assignment, and remove quotes (if they exist)
+        auto delim = assignment.find('=');
+        if (delim == std::string::npos)
+            throw std::invalid_argument("EDFHandler::read: failed reading name/value delimiter");
+        std::string name = assignment.substr(0, delim);
+        StringUtils::trim(name);
+        std::string value = assignment.substr(delim+1);
+        StringUtils::trim(value);
+        if (value[0] == value[value.size()-1] && (value[0] == '\'' || value[0] == '\"'))
+            value = value.substr(1, value.size() - 2);
+
+        StringUtils::tolower(name);
+        attributes[name] = value;
+    }
+    file.ignore(2); // end of header marker
+
+    // --- now parse the header data
+
+    // read the dimensions
+    std::vector<int> dim;
+    for (int i = 1; ; ++i) {
+        // assemble the attribute name
+        std::stringstream aux;
+        aux << "dim_" << i;
+
+        // try to find the attribute
+        auto dimIt = attributes.find(aux.str());
+        if (dimIt == attributes.end())
+            break;
+
+        dim.push_back(StringUtils::parseString<int>(dimIt->second));
+    }
+    const int nDims = dim.size();
+    if (!nDims)
+        throw std::runtime_error("EDFHandler::read: dimension information not found");
+
+    // parse the (non-standard) spacing tag
+    std::vector<float> spacing;
+    auto spacingIt = attributes.find("spacing");
+    if (spacingIt != attributes.end())
+        StringUtils::parseString<float>(spacingIt->second, spacing);
+
+    // check for a byte order tag, but fall back to the default value
+    ByteOrder::ByteOrderEnum byteorder = ByteOrder::DEFAULT;
+    auto byteorderIt = attributes.find("byteorder");
+    if (byteorderIt != attributes.end()) {
+        std::string byteorderValue = byteorderIt->second;
+        StringUtils::tolower(byteorderValue);
+
+        if (byteorderValue == "highbytefirst")
+            byteorder = ByteOrder::HIGH_BYTE_FIRST;
+        else if (byteorderValue == "lowbytefirst")
+            byteorder = ByteOrder::LOW_BYTE_FIRST;
+        else
+            throw std::runtime_error("invalid byte order value");
+    }
+
+    // check for the 'element type' value
+    DataTypes::DataTypeEnum datatype;
+    auto datatypeIt = attributes.find("datatype");
+    if (datatypeIt != attributes.end()) {
+        std::string datatypeValue = datatypeIt->second;
+        StringUtils::tolower(datatypeValue);
+
+        if (datatypeValue == "signedbyte")
+            datatype = DataTypes::INT8;
+        else if (datatypeValue == "unsignedbyte")
+            datatype = DataTypes::UINT8;
+        else if (datatypeValue == "signedshort")
+            datatype = DataTypes::INT16;
+        else if (datatypeValue == "unsignedshort")
+            datatype = DataTypes::UINT16;
+        else if (datatypeValue == "float" || datatypeValue == "floatvalue" || datatypeValue == "real")
+            datatype = DataTypes::FLOAT32;
+        else if (datatypeValue == "double" || datatypeValue == "doublevalue")
+            datatype = DataTypes::FLOAT64;
+        // todo: add further data types
+        else
+            throw std::runtime_error("EDFHandler::read: invalid/unsupported data type");
+    }
+    else
+        throw std::runtime_error("EDFHandler::read: data type not found");
+
+    auto compressionIt = attributes.find("compression");
+    if (compressionIt != attributes.end())
+        throw std::runtime_error("compression not supported");
+
+    int size = 0;
+    auto sizeIt = attributes.find("size");
+    if (sizeIt != attributes.end())
+        size = StringUtils::parseString<int>(sizeIt->second);
+
+    auto imageIt = attributes.find("image");
+    if (imageIt != attributes.end() && StringUtils::parseString<int>(imageIt->second) != 1)
+        throw std::runtime_error("EDFHandler::read: image not set to 1");
+
+    // convert size
+    Eigen::VectorXi dimSizes(nDims);
+    for (int i = 0; i < nDims; ++i)
+        dimSizes[i] = dim[i];
+    if (dimSizes.prod() * DataTypes::getSizeForType(datatype) != size)
+        throw std::runtime_error("EDFHandler::read: size inconsistency");
+
+    // convert spacing
+    Eigen::VectorXf dimSpacing = Eigen::VectorXf::Ones(nDims);
+    if (spacing.size() > 0) {
+        if (spacing.size() != nDims)
+            throw std::runtime_error("EDFHandler::read: spacing inconsistency");
+        for (int i = 0; i < nDims; ++i)
+            dimSpacing[i] = spacing[i];
+    }
+
+    // compute the byte shuffling flag
+    const bool byteShuffle = (byteorder != ByteOrder::DEFAULT);
+
+
+    // --- read in the image data
+    Eigen::VectorXf data = DataTypes::parseData<float>(file, datatype, dimSizes.prod(), byteShuffle);
+
+    file.close();
+
+
+    // --- setup the volume
+    Eigen::Vector3f lowerLeft = Eigen::Vector3f::Zero();
+    Eigen::Vector3f upperRight = lowerLeft + (dimSizes.cast<float>().array() * dimSpacing.array() ).matrix();
+
+    Volume vol(lowerLeft, upperRight, dimSpacing);
+    vol.setContent(data);
+
+    return vol;
+}
+
+
+void EDFHandler::write(const std::string& filename, const Volume& vol)
+{
+    // open the EDF file
+    std::ofstream file(filename, std::ios::binary);
+    if (!file.good())
+        throw std::invalid_argument("EDFHandler::write: cannot open " + filename + " for writing");
+
+    // open the header
+    file << "{\n";
+
+    file << "HeaderID = EH:000001:000000:000000;\n";
+    file << "Image = " << 1 << ";\n";
+    if (ByteOrder::DEFAULT == ByteOrder::HIGH_BYTE_FIRST)
+        file << "ByteOrder = HighByteFirst;\n";
+    else
+        file << "ByteOrder = LowByteFirst;\n";
+    file << "DataType = " << TYPE_FLOAT << ";\n";
+
+    // write dimension and size
+    for (int i = 0; i < 3; ++i)
+        file << "Dim_" << (i+1) << " = " << vol.getNumVoxels()[i] << ";\n";
+    file << "Size = " << vol.getNumVoxels().prod() * sizeof(float) << ";\n";
+
+    // write spacing
+    file << "Spacing =";
+    for (int i = 0; i < 3; ++i)
+        file << ' ' << vol.getSpacing()[i];
+    file << ";\n";
+
+    // pad the header by adding spaces such that the header ends on a kilobyte boundary
+    std::size_t n = 1024;
+    while (n < (static_cast<std::size_t>(file.tellp()) + 3))
+        n += 1024;
+    n -= static_cast<std::size_t>(file.tellp()) + 3;
+    while (n > 0) {
+        file.put(' ');
+        n--;
+    }
+
+    // close the header
+    file << "\n}\n";
+
+    // write the volume data
+    file.write(reinterpret_cast<const char*>(vol.getContent().data()), vol.getNumVoxels().prod() * sizeof(float) );
+    file.close();
+}
+

modules/cppcourse/processors/EDFhandler.h

+#ifndef EDFHANDLER_H
+#define EDFHANDLER_H
+
+#include <string>
+
+#include "volume.h"
+
+
+/**
+ * \brief EDFHandler reads and writes EDF files.
+ */
+
+class EDFHandler {
+public:
+    /* \brief read an EDF file into a Volume
+     * \param[in] filename - filename of the EDF file to be read
+     * \return volume containing the data
+     */
+    static Volume read(const std::string& filename);
+
+    /* \brief write out a Volume into an EDF file
+     * \param[in] filename - filename of the EDF file to be written
+     * \param[in] vol      - volume to be written
+     */
+    static void write(const std::string& filename, const Volume& vol);
+
+private:
+    static constexpr auto TYPE_FLOAT = "FloatValue";
+};
+
+
+#endif // EDFHANDLER_H
+

modules/cppcourse/processors/IOutilities.h

+#ifndef IOUTILITIES_H
+#define IOUTILITIES_H
+
+#include <algorithm>
+#include <regex>
+#include <string>
+#include <vector>
+
+#include <Eigen/Dense>
+
+
+/**
+ * \brief StringUtils provides common string utilities for file parsing.
+ */
+
+struct StringUtils {
+        /* \brief remove whitespace at front and back of string
+         * \param[in,out] str - string to be trimmed
+         */
+        static void trim(std::string& str);
+
+        /* \brief convert string to lower case
+         * \param[in,out] str - string to be converted
+         */
+        static void tolower(std::string& str);
+
+        /* \brief convert string to upper case
+         * \param[in,out] str - string to be converted
+         */
+        static void toupper(std::string& str);
+
+        /* \brief convert string to arbitrary datatype using stringstream
+         * \param[in] str - string to be converted
+         * \return converted datatype
+         */
+        template<typename Type>
+        static Type parseString(const std::string& str);
+
+        /* \brief convert string to vector of arbitrary data type using stringstream
+         * \param[in] str - string to be converted
+         * \param[out] vec - vector of converted datatypes
+         * \return number of converted datatypes
+         */
+        template<typename Type>
+        static int parseString(const std::string& str, std::vector<Type>& vec);
+};
+
+
+
+/**
+ * \brief ByteOrder supplies helper routines to cope with byte order conversions.
+ */
+
+struct ByteOrder {
+    /// byte order enum
+    enum ByteOrderEnum {
+        LOW_BYTE_FIRST,     // little endian (e.g. x86)
+        HIGH_BYTE_FIRST     // big endian (e.g. PowerPC)
+    };
+
+    /// attempt to determine endianness (works on Mac at least..), default to little endian
+    static const ByteOrderEnum DEFAULT =
+#if defined(BYTE_ORDER) && defined(BIG_ENDIAN) && BYTE_ORDER == BIG_ENDIAN
+            HIGH_BYTE_FIRST;
+#else
+            LOW_BYTE_FIRST;
+#endif
+
+    /* \brief shuffle the bytes of the data pointed to by ptr to switch endianness
+     * \tparam _DataType       - data type underlying the pointer
+     * \param[in,out] ptr      - pointer to the data
+     * \param[in]     elements - number of elements of _DataType
+     */
+    template<typename _DataType>
+    static void shuffleBytes(void* ptr, long elements);
+};
+
+
+
+/**
+ * \brief DataTypes supplies helper enums and routines for type conversions.
+ */
+
+struct DataTypes {
+    /// data type enum
+    enum DataTypeEnum {
+        INT8,  UINT8,
+        INT16, UINT16,
+        INT32, UINT32,
+        FLOAT32,
+        FLOAT64
+    };
+
+    /* \brief return the size in bytes for the specified data type
+     * \param[in] dataType - type for which to return the size in bytes
+     * \return size of dataType in bytes
+     */
+    static int getSizeForType(const DataTypeEnum& dataType);
+
+    /* \brief parse data of type dataType from stream in into an Eigen::Vector
+     *        (dispatches to parseData<_DataType, _Scalar> internally)
+     * \tparam _Scalar           - scalar type of Eigen::Vector to return
+     * \param[in] in             - stream from which to read
+     * \param[in] dataType       - dataType of elements to be read from stream
+     * \param[in] sizeInElements - number of elements (of type dataType) to read from stream
+     * \param[in] shuffleBytes   - flag whether to shuffle bytes around to account for endianness change
+     * \return vector of _Scalar containing the read in values
+     */
+    template<typename _Scalar>
+    static Eigen::Matrix<_Scalar, Eigen::Dynamic, 1> parseData(std::istream& in, const DataTypeEnum& dataType, long sizeInElements, bool shuffleBytes);
+
+private:
+    /* \brief parse data of type _DataType from stream in into an Eigen::Vector
+     * \tparam _DataType         - data type of elements to be read from stream
+     * \tparam _Scalar           - scalar type of Eigen::Vector to return
+     * \param[in] in             - stream from which to read
+     * \param[in] sizeInElements - number of elements (of type _DataType) to read from stream
+     * \param[in] shuffleBytes   - flag whether to shuffle bytes around to account for endianness change
+     * \return vector of _Scalar containing the read in values
+     */
+    template<typename _DataType, typename _Scalar>
+    static Eigen::Matrix<_Scalar, Eigen::Dynamic, 1> parseData(std::istream& in, long sizeInElements, bool shuffleBytes);
+};
+
+
+
+
+
+// --------- Implementations -----------
+
+inline void StringUtils::trim(std::string& str)
+{
+    str = std::regex_replace(str, std::regex("^\\s+|\\s+$"), std::string(""));
+}
+
+
+inline void StringUtils::tolower(std::string& str)
+{
+    std::transform(str.begin(), str.end(), str.begin(), ::tolower);
+}
+
+
+inline void StringUtils::toupper(std::string& str)
+{
+    std::transform(str.begin(), str.end(), str.begin(), ::toupper);
+}
+
+
+template<typename Type>
+inline Type StringUtils::parseString(const std::string& str)
+{
+    Type value;
+    std::stringstream convert(str);