Project 3: Xuecheng Sun

CMakeLists.txt

@@ -71,8 +71,15 @@ set(headers
     src/pathtrace.h
     src/scene.h
     src/sceneStructs.h
+    src/json.hpp
     src/preview.h
     src/utilities.h
+    src/tiny_gltf.h
+    src/macros.h
+    src/gltf-loader.h
+    src/material.h
+    src/mesh.h
+    src/octree.h
     )
 
 set(sources
@@ -84,6 +91,8 @@ set(sources
     src/scene.cpp
     src/preview.cpp
     src/utilities.cpp
+    src/gltf-loader.cc
+    src/octree.cpp
     )
 
 list(SORT headers)
@@ -92,7 +101,7 @@ list(SORT sources)
 source_group(Headers FILES ${headers})
 source_group(Sources FILES ${sources})
 
-#add_subdirectory(stream_compaction)  # TODO: uncomment if using your stream compaction
+add_subdirectory(stream_compaction)  # TODO: uncomment if using your stream compaction
 
 cuda_add_executable(${CMAKE_PROJECT_NAME} ${sources} ${headers})
 target_link_libraries(${CMAKE_PROJECT_NAME}

README.md

@@ -3,11 +3,83 @@ CUDA Path Tracer
 
 **University of Pennsylvania, CIS 565: GPU Programming and Architecture, Project 3**
 
-* (TODO) YOUR NAME HERE
-* Tested on: (TODO) Windows 22, i7-2222 @ 2.22GHz 22GB, GTX 222 222MB (Moore 2222 Lab)
+* Xuecheng Sun
+* Tested on: Windows 10, R7 3700x @3.7GHz, RTX 2070 super 8GB
 
-### (TODO: Your README)
 
-*DO NOT* leave the README to the last minute! It is a crucial part of the
-project, and we will not be able to grade you without a good README.
 
+### Customized Scene: The Arsenal
+
+![](./img/Arsenel_Blue.png)
+
+![](./img/Arsenel_White.png)
+
+### Path Tracer Part 1
+
+**Finished Feature:** 
+
+	1. Specular and Diffuse Ray Scatter
+ 	2. Basic stream compaction of stop thread
+ 	3. First bounce caching
+
+![](./img/PathTracerPart1.png)
+
+
+
+**Benchmark for First Ray Cache**
+
+1. Sorted material can reduce the divergence in a thread warp which can significantly boost the render speed.
+
+2. First bounce cache always provide a constant value improvement with different depth setting
+
+![](img/PathTracerDepth.png)
+
+### Path Tracer Part 2 and 3
+
+**1.  Different Materials**
+
+Implemented perfect refractive, perfect reflective and Schlick's approximation fresnel materials:
+
+![](./img/DiffMat.png)
+
+**2. Depth of Field and AA**
+
+![](./img/DiffMat_DOF_AA_Comp.png)
+
+**3.Arbitrary Mesh(GLTF) Import**
+
+​	**Extra Points: Texture Mapping: Base Color, Normal and Roughness**
+
+![](./img/GLTF.png)
+
+**4. Direct Light**
+
+![](./img/directLighting.png)
+
+**5. Advance Sampling Method: Halton Sequence**
+
+**Halton Direct Light Sampling**
+
+![](./img/HaltonDirect.png)
+
+**Random Direct Light Sampling as Reference**
+
+![](./img/RandomDirect.png)
+
+Shadows under Halton Sampling are more regularized.
+
+**6. Octree(Partially)**
+
+I build the octree for primitives and triangle but find for mesh triangles, it always aggregates in the root node which finally slow down the rendering. 
+
+### Benchmark for Bounding Box
+
+![](./img/BoundingBoxBM.png)
+
+For a very simple mesh such as cube the bounding box didn't help much for the rendering speed, but if we render a more complex mesh like pistol or duck which have more than 4000 triangles, the bounding box improved the performance greatly.
+
+
+
+### Third Libraries
+
+Tinygltf libraries: GLTF Loader

build/ALL_BUILD.vcxproj.filters

@@ -0,0 +1,8 @@
+<?xml version="1.0" encoding="utf-8"?>
+<Project ToolsVersion="16.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
+  <ItemGroup>
+    <CustomBuild Include="E:\Learning\CIS565\Project3-CUDA-Path-Tracer\CMakeLists.txt" />
+  </ItemGroup>
+  <ItemGroup>
+  </ItemGroup>
+</Project>

build/ALL_BUILD.vcxproj.user

@@ -0,0 +1,4 @@
+<?xml version="1.0" encoding="utf-8"?>
+<Project ToolsVersion="Current" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
+  <PropertyGroup />
+</Project>