当前位置：首页 > news >正文

【OpenGL】LearnOpenGL学习笔记26 - 视差贴图 Parallax Map

news 2025/9/22 10:00:54

上接：https://blog.csdn.net/weixin_44506615/article/details/151898818?spm=1001.2014.3001.5501
完整代码：https://gitee.com/Duo1J/learn-open-gl | https://github.com/Duo1J/LearnOpenGL

视差贴图 (Parallax Map)

视差贴图类似法线贴图，它利用了视错觉，能够极大地提升表面细节，使之具有深度感
视差贴图背后的思想是通过修改 纹理坐标(UV) 来使一个片段的表面看起来比实际更高或是更低，如下图所示 (图片来自于LearnOpenGL)

如果不进行干涉，观察者的视线就会落到点A上，我们的目的就是让A上的片段不再使用点A的纹理坐标，而是使用点B的纹理坐标进行采样，这样观察者的视线就仿佛落在了点B

首先准备一下要用到的纹理图片，我们绘制两种平面来看效果，砖墙和玩具盒，相关资源可在顶部Git仓库中找到
视差贴图
首先绘制两个平面，并应用法线贴图
Main.cpp

// ...
// 渲染视差Quad
void RenderParallax()
{static unsigned int parallaxVAO = 0;static unsigned int parallaxVBO;if (parallaxVAO == 0){// positionsglm::vec3 pos1(-1.0, 1.0, 0.0);glm::vec3 pos2(-1.0, -1.0, 0.0);glm::vec3 pos3(1.0, -1.0, 0.0);glm::vec3 pos4(1.0, 1.0, 0.0);// texture coordinatesglm::vec2 uv1(0.0, 1.0);glm::vec2 uv2(0.0, 0.0);glm::vec2 uv3(1.0, 0.0);glm::vec2 uv4(1.0, 1.0);// normalglm::vec3 nm(0.0, 0.0, 1.0);glm::vec3 tangent1, bitangent1;glm::vec3 tangent2, bitangent2;// - triangle 1glm::vec3 edge1 = pos2 - pos1;glm::vec3 edge2 = pos3 - pos1;glm::vec2 deltaUV1 = uv2 - uv1;glm::vec2 deltaUV2 = uv3 - uv1;GLfloat f = 1.0f / (deltaUV1.x * deltaUV2.y - deltaUV2.x * deltaUV1.y);tangent1.x = f * (deltaUV2.y * edge1.x - deltaUV1.y * edge2.x);tangent1.y = f * (deltaUV2.y * edge1.y - deltaUV1.y * edge2.y);tangent1.z = f * (deltaUV2.y * edge1.z - deltaUV1.y * edge2.z);tangent1 = glm::normalize(tangent1);bitangent1.x = f * (-deltaUV2.x * edge1.x + deltaUV1.x * edge2.x);bitangent1.y = f * (-deltaUV2.x * edge1.y + deltaUV1.x * edge2.y);bitangent1.z = f * (-deltaUV2.x * edge1.z + deltaUV1.x * edge2.z);bitangent1 = glm::normalize(bitangent1);// - triangle 2edge1 = pos3 - pos1;edge2 = pos4 - pos1;deltaUV1 = uv3 - uv1;deltaUV2 = uv4 - uv1;f = 1.0f / (deltaUV1.x * deltaUV2.y - deltaUV2.x * deltaUV1.y);tangent2.x = f * (deltaUV2.y * edge1.x - deltaUV1.y * edge2.x);tangent2.y = f * (deltaUV2.y * edge1.y - deltaUV1.y * edge2.y);tangent2.z = f * (deltaUV2.y * edge1.z - deltaUV1.y * edge2.z);tangent2 = glm::normalize(tangent2);bitangent2.x = f * (-deltaUV2.x * edge1.x + deltaUV1.x * edge2.x);bitangent2.y = f * (-deltaUV2.x * edge1.y + deltaUV1.x * edge2.y);bitangent2.z = f * (-deltaUV2.x * edge1.z + deltaUV1.x * edge2.z);bitangent2 = glm::normalize(bitangent2);GLfloat quadVertices[] = {// Positions            // normal         // TexCoords  // Tangent                          // Bitangentpos1.x, pos1.y, pos1.z, nm.x, nm.y, nm.z, uv1.x, uv1.y, tangent1.x, tangent1.y, tangent1.z, bitangent1.x, bitangent1.y, bitangent1.z,pos2.x, pos2.y, pos2.z, nm.x, nm.y, nm.z, uv2.x, uv2.y, tangent1.x, tangent1.y, tangent1.z, bitangent1.x, bitangent1.y, bitangent1.z,pos3.x, pos3.y, pos3.z, nm.x, nm.y, nm.z, uv3.x, uv3.y, tangent1.x, tangent1.y, tangent1.z, bitangent1.x, bitangent1.y, bitangent1.z,pos1.x, pos1.y, pos1.z, nm.x, nm.y, nm.z, uv1.x, uv1.y, tangent2.x, tangent2.y, tangent2.z, bitangent2.x, bitangent2.y, bitangent2.z,pos3.x, pos3.y, pos3.z, nm.x, nm.y, nm.z, uv3.x, uv3.y, tangent2.x, tangent2.y, tangent2.z, bitangent2.x, bitangent2.y, bitangent2.z,pos4.x, pos4.y, pos4.z, nm.x, nm.y, nm.z, uv4.x, uv4.y, tangent2.x, tangent2.y, tangent2.z, bitangent2.x, bitangent2.y, bitangent2.z};glGenVertexArrays(1, &parallaxVAO);glGenBuffers(1, &parallaxVBO);glBindVertexArray(parallaxVAO);glBindBuffer(GL_ARRAY_BUFFER, parallaxVBO);glBufferData(GL_ARRAY_BUFFER, sizeof(quadVertices), &quadVertices, GL_STATIC_DRAW);glEnableVertexAttribArray(0);glVertexAttribPointer(0, 3, GL_FLOAT, GL_FALSE, 14 * sizeof(GLfloat), (GLvoid*)0);glEnableVertexAttribArray(1);glVertexAttribPointer(1, 3, GL_FLOAT, GL_FALSE, 14 * sizeof(GLfloat), (GLvoid*)(3 * sizeof(GLfloat)));glEnableVertexAttribArray(2);glVertexAttribPointer(2, 2, GL_FLOAT, GL_FALSE, 14 * sizeof(GLfloat), (GLvoid*)(6 * sizeof(GLfloat)));glEnableVertexAttribArray(3);glVertexAttribPointer(3, 3, GL_FLOAT, GL_FALSE, 14 * sizeof(GLfloat), (GLvoid*)(8 * sizeof(GLfloat)));glEnableVertexAttribArray(4);glVertexAttribPointer(4, 3, GL_FLOAT, GL_FALSE, 14 * sizeof(GLfloat), (GLvoid*)(11 * sizeof(GLfloat)));}glBindVertexArray(parallaxVAO);glDrawArrays(GL_TRIANGLES, 0, 6);glBindVertexArray(0);
}int main()
{// ...Shader parallarShader("Shader/ParallaxVertex.glsl", "Shader/ParallaxFragment.glsl");// ...Texture bricksDiffuseTex("F:/Scripts/Cpp/LearnOpenGL/learn-open-gl/Resource/parallax/T_Bricks_Diffuse.jpg", TextureType::DIFFUSE, GL_CLAMP_TO_EDGE);Texture bricksNormalTex("F:/Scripts/Cpp/LearnOpenGL/learn-open-gl/Resource/parallax/T_Bricks_Normal.jpg", TextureType::NORMAL, GL_CLAMP_TO_EDGE);Texture bricksParallaxTex("F:/Scripts/Cpp/LearnOpenGL/learn-open-gl/Resource/parallax/T_Bricks_Parallax.jpg", TextureType::DIFFUSE, GL_CLAMP_TO_EDGE);Texture toyboxDiffuseTex("F:/Scripts/Cpp/LearnOpenGL/learn-open-gl/Resource/parallax/T_Toybox_Diffuse.png", TextureType::DIFFUSE, GL_CLAMP_TO_EDGE);Texture toyboxNormalTex("F:/Scripts/Cpp/LearnOpenGL/learn-open-gl/Resource/parallax/T_Toybox_Normal.png", TextureType::NORMAL, GL_CLAMP_TO_EDGE);Texture toyboxParallaxTex("F:/Scripts/Cpp/LearnOpenGL/learn-open-gl/Resource/parallax/T_Toybox_Parallax.png", TextureType::DIFFUSE, GL_CLAMP_TO_EDGE);// [主循环]// 绘制视差QuadparallarShader.Use();parallarShader.SetMat4("view", view);parallarShader.SetMat4("projection", projection);glm::mat4 parallaxModel(1);parallaxModel = glm::translate(parallaxModel, glm::vec3(0, -3, 2));parallarShader.SetMat4("model", parallaxModel);parallarShader.SetVec3("lightPos", glm::vec3(1.7f, -2.8f, 5.0f));parallarShader.SetVec3("viewPos", camera.transform.position);parallarShader.SetInt("diffuseMap", 0);parallarShader.SetInt("normalMap", 1);parallarShader.SetInt("parallaxMap", 2);glDisable(GL_CULL_FACE);// 砖墙glActiveTexture(GL_TEXTURE0);glBindTexture(GL_TEXTURE_2D, bricksDiffuseTex.GetTextureID());glActiveTexture(GL_TEXTURE1);glBindTexture(GL_TEXTURE_2D, bricksNormalTex.GetTextureID());glActiveTexture(GL_TEXTURE2);glBindTexture(GL_TEXTURE_2D, bricksParallaxTex.GetTextureID());RenderParallax();//玩具盒glActiveTexture(GL_TEXTURE0);glBindTexture(GL_TEXTURE_2D, toyboxDiffuseTex.GetTextureID());glActiveTexture(GL_TEXTURE1);glBindTexture(GL_TEXTURE_2D, toyboxNormalTex.GetTextureID());glActiveTexture(GL_TEXTURE2);glBindTexture(GL_TEXTURE_2D, toyboxParallaxTex.GetTextureID());glDisable(GL_CULL_FACE);parallaxModel = glm::translate(parallaxModel, glm::vec3(-3, 0, 0));parallarShader.SetMat4("model", parallaxModel);RenderParallax();glEnable(GL_CULL_FACE);
}

ParallaxVertex.glsl 新建

#version 330 corelayout (location = 0) in vec3 position;
layout (location = 1) in vec3 normal;
layout (location = 2) in vec2 texCoords;
layout (location = 3) in vec3 tangent;
layout (location = 4) in vec3 bitangent;out VS_OUT {vec3 FragPos;vec2 TexCoords;vec3 TangentLightPos;vec3 TangentViewPos;vec3 TangentFragPos;
} vs_out;uniform mat4 projection;
uniform mat4 view;
uniform mat4 model;uniform vec3 lightPos;
uniform vec3 viewPos;void main()
{gl_Position = projection * view * model * vec4(position, 1.0f);vs_out.FragPos = vec3(model * vec4(position, 1.0));   vs_out.TexCoords = texCoords;vec3 T = normalize(mat3(model) * tangent);vec3 B = normalize(mat3(model) * bitangent);vec3 N = normalize(mat3(model) * normal);mat3 TBN = transpose(mat3(T, B, N));vs_out.TangentLightPos = TBN * lightPos;vs_out.TangentViewPos  = TBN * viewPos;vs_out.TangentFragPos  = TBN * vs_out.FragPos;
}

ParallaxFragment.glsl 新建

#version 330 coreout vec4 FragColor;in VS_OUT {vec3 FragPos;vec2 TexCoords;vec3 TangentLightPos;vec3 TangentViewPos;vec3 TangentFragPos;
} fs_in;uniform sampler2D diffuseMap;
uniform sampler2D normalMap;
uniform sampler2D depthMap;void main()
{           vec3 viewDir = normalize(fs_in.TangentViewPos - fs_in.TangentFragPos);vec2 texCoords = fs_in.TexCoords;vec3 normal = texture(normalMap, texCoords).rgb;normal = normalize(normal * 2.0 - 1.0);   vec3 color = texture(diffuseMap, texCoords).rgb;// Ambientvec3 ambient = 0.1 * color;// Diffusevec3 lightDir = normalize(fs_in.TangentLightPos - fs_in.TangentFragPos);float diff = max(dot(lightDir, normal), 0.0);vec3 diffuse = diff * color;// Specular    vec3 reflectDir = reflect(-lightDir, normal);vec3 halfwayDir = normalize(lightDir + viewDir);  float spec = pow(max(dot(normal, halfwayDir), 0.0), 32.0);vec3 specular = vec3(0.2) * spec;FragColor = vec4(ambient + diffuse + specular, 1.0f);
}

编译运行
视差贴图

视差映射 (Parallax Mapping)

ParallaxFragment.glsl

uniform float heightScale;// 视差映射
vec2 ParallaxMapping(vec2 texCoords, vec3 viewDir)
{float depth = texture(parallaxMap, texCoords).r;vec2 p = viewDir.xy / viewDir.z * (depth * depthScale);return texCoords - p;
}void main()
{           vec3 viewDir = normalize(fs_in.TangentViewPos - fs_in.TangentFragPos);vec2 texCoords = ParallaxMapping(fs_in.TexCoords, viewDir);// ...
}

定义一个名为 ParallaxMapping 的函数，它接收UV坐标和视线方向
首先使用原始的UV坐标采样视差贴图获取深度值 depth
然后 viewDir.xy / viewDir.z 进行透视矫正
最后乘以采样来的深度值 depth 和外部传入的常量 depthScale 获取UV偏移量
Main.cpp

parallarShader.SetFloat("depthScale", 0.1f);

编译运行，顺利的话可以看见以下图像
视差映射

砖墙的边缘看起来还有点怪，这是因为纹理坐标经过偏移超出了 [0, 1] 的范围，我们需要进行剔除
ParallaxFragment.glsl

if(texCoords.x > 1.0 || texCoords.y > 1.0 || texCoords.x < 0.0 || texCoords.y < 0.0)discard;

视差映射剔除

三、陡峭视差映射 (Steep Parallax Mapping)

如果以大角度的侧面进行观察会发现一些问题
视察映射走样
这是因为我们目前只是进行了一个大致的视差映射，接下来使用陡峭视差映射来获取更精确的结果

陡峭视差映射的基本思想是将总深度范围划分为同一个深度/高度的多个层，我们从上至下遍历深度层，如果该层的深度值高于了存储在深度贴图中的值则结束

// 陡峭视差映射
vec2 SteepParallaxMapping(vec2 texCoords, vec3 viewDir)
{// 划分的层数const float numLayers = 10;// 每层大小float layerDepth = 1.0 / numLayers;// 当前层深度float currentLayerDepth = 0.0;// 纹理坐标变化量vec2 P = viewDir.xy * depthScale; vec2 deltaTexCoords = P / numLayers;vec2  currentTexCoords     = texCoords;float currentDepthMapValue = texture(parallaxMap, currentTexCoords).r;while (currentLayerDepth < currentDepthMapValue){currentTexCoords -= deltaTexCoords;currentDepthMapValue = texture(parallaxMap, currentTexCoords).r;  currentLayerDepth += layerDepth;  }return currentTexCoords;
}

编译运行，从侧面看效果好了很多
陡峭视差映射
不过由于采样深度变化非连续，会出现断层问题

可以通过观察的角度来优化层数，倾斜角度越大层数越多，提供更好的性能的同时减轻断层的问题

const float minLayers = 8;
const float maxLayers = 32;float numLayers = mix(maxLayers, minLayers, abs(dot(vec3(0.0, 0.0, 1.0), viewDir)));

但断层的根本原因是因为采样深度变化非连续，接下来我们使用视差遮蔽映射来解决这个问题

四、视差遮蔽映射 (Parallax Occlusion Mapping)

视差遮蔽映射和陡峭视差映射类似，但不是直接使用目标层的纹理坐标，而是根据表面高度距离来在深度层之间进行线性插值

ParallaxFragment.glsl

// 视差遮蔽映射
vec2 ParallaxOcclusionMapping(vec2 texCoords, vec3 viewDir)
{ // 划分的层数const float minLayers = 10;const float maxLayers = 20;float numLayers = mix(maxLayers, minLayers, abs(dot(vec3(0.0, 0.0, 1.0), viewDir)));  // 每层大小float layerDepth = 1.0 / numLayers;// 当前层深度float currentLayerDepth = 0.0;// 纹理坐标变化量vec2 P = viewDir.xy / viewDir.z * depthScale; vec2 deltaTexCoords = P / numLayers;vec2  currentTexCoords     = texCoords;float currentDepthMapValue = texture(parallaxMap, currentTexCoords).r;while(currentLayerDepth < currentDepthMapValue){currentTexCoords -= deltaTexCoords;currentDepthMapValue = texture(parallaxMap, currentTexCoords).r;  currentLayerDepth += layerDepth;  }// 上一个深度的纹理坐标vec2 prevTexCoords = currentTexCoords + deltaTexCoords;// 获取两层的深度，用于后续插值float afterDepth  = currentDepthMapValue - currentLayerDepth;float beforeDepth = texture(parallaxMap, prevTexCoords).r - currentLayerDepth + layerDepth;// 插值权重float weight = afterDepth / (afterDepth - beforeDepth);vec2 finalTexCoords = prevTexCoords * weight + currentTexCoords * (1.0 - weight);return finalTexCoords;
}