理论知识转载地址:http://blog.csdn.net/ym19860303/article/details/25545933

1.Lambert模型(漫反射)

环境光:

Iambdiff = Kd*Ia

其中Ia 表示环境光强度,Kd(0<K<1)为材质对环境光的反射系数,Iambdiff是漫反射体与环境光交互反射的光强。

方向光:

Ildiff = Kd * Il * Cos(θ)

其中Il是点光源强度,θ是入射光方向与顶点法线的夹角,称入射角(0<=A<=90°),Ildiff是漫反射体与方向光交互反射的光强,若 N为顶点单位法向量,L表示从顶点指向光源的单位向量(注意顶点指向光源),则Cos(θ)等价于dot(N,L),故又有:

Ildiff = Kd * Il * dot(N,L)

最后综合环境光和方向光源,Lambert光照模型可以写成:

Idiff = Iambdiff + Ildiff = Kd * Ia + Kd * Il * dot(N,L)

2.Phong模型(镜面反射)

Phong模型认为镜面反射的光强与反射光线和视线的夹角相关:

Ispec = Ks * Il * ( dot(V,R) )^Ns

其中Ks 为镜面反射系数,Ns是高光指数,V表示从顶点到视点的观察方向,R代表反射光方向。由于反射光的方向R可以通过入射光方向L(从顶点指向光源)和物体的法向量求出,
R + L = 2 * dot(N, L) * N  即 R = 2 * dot(N,L) * N - L

所以最终的计算式为:

Ispec = Ks * Il * ( dot(V, (2 * dot(N,L) * N – L ) )^Ns

3.Blinn-Phong光照模型(修正镜面光)

Blinn-Phong是一个基于Phong模型修正的模型,其公式为:

Ispec = Ks * Il * ( dot(N,H) )^Ns

其中N是入射点的单位法向量,H是光入射方向L和视点方向V的中间向量,通常也称之为半角向量(半角向量被广泛用于各类光照模型,原因不但在于半角向量蕴含的信息价值,也在于半角向量是很简单的计算:H = (L + V) / |L + V|  )。

4.Rendering Equation(全局光照模型)

Rendering Equation 是Kajia在1986年提出的,

Lo(X, Wo) = Le(X, Wo) + ∫fr(X, Wi, Wo) Li(X, Wi) dot(N, Wi) dWi

其中X表示入射点,Lo(X, Wo)即从物体表面X点,沿方向Wo反射的光强,Le(X, Wo)表示从物体表面X以方向Wo 发射出去的光强,该值仅对自发光体有效,fr(X, Wi, Wo)为,入射光线方向为Wi, 照射到点X上,然后从Wo方向发射出去的BRDF值,Li(X, Wi)为入射方向为Wi照射到点X上的入射光强,N表示点X处的法向量,然后对入射方向进行积分(因为光线入射的方向是四面八方的,积分的意义是对每个方向进行一遍计算后相加),计算的结果就是全局光照的辐射率。

对于单个点光源照射到不会自发光的物体上,公式可以简化成:

Lo(X, Wo) = fr(X, Wi, Wo) Li(X, Wi) dot(N, Wi)

这个公式非常有用,通常会将该公式分解为漫反射表达式和镜面表达式之和。对于漫反射表面,BRDF可以忽略不计,因为它总是返回某个恒定值,所以可以写成如下形式:

Lo(X, Wo) = Idiff + frs(X, Wi, Wo) Li(X, Wi) dot(N, Wi)

其中Idiff表示漫反射分量,使用公式的计算方法,frs(X, Wi, Wo)表示镜面反射的BRDF函数,前面的Phong高光模型,其实是rendering equation在单一光源下针对理想镜面反射的特定推导,对于Phong高光而言:

frs(X, Wi, Wo) = Ks (dot(N, H)^Ns  / dot(N, Wi)

基于GLSL的实现(所有光照和材质参数在实际应用时,应该从常量改为uniform,由外部进行控制):

// 在顶点着色器中处理光照计算(环境光+漫射光+镜面反射)
const char* ccPositionTextureColorForLight1_v = " \
uniform vec2 translate; \n\
\n\
attribute vec4 a_position; \n\
attribute vec4 a_color; \n\
attribute vec2 a_texCoord; \n\
attribute vec3 a_normal; \n\
\n\
#ifdef GL_ES \n\
varying mediump vec4 v_color; \n\
varying mediump vec2 v_texCoord; \n\
#else \n\
varying vec4 v_color; \n\
varying vec2 v_texCoord; \n\
#endif \n\
\n\
const vec3 lightAmbient = vec3(1.0, 1.0, 1.0); \n\
const vec3 lightDiffuse = vec3(1.0, 1.0, 1.0); \n\
const vec3 lightSpecular = vec3(1.0, 1.0, 1.0); \n\
const vec3 materialAmbient = vec3(1.0, 1.0, 1.0); \n\
const vec3 materialDiffuse = vec3(1.0, 1.0, 1.0); \n\
const vec3 materialSpecular = vec3(1.0, 1.0, 1.0); \n\
const int shininess = 8; // 0-128 \n\
\n\
const vec3 eyePosition = vec3(512.0, 128.0, 442.0); \n\
const vec3 lightDirection = vec3(1.0, 1.0, 1.0); \n\
\n\
void main() { \n\
vec4 position = CC_MVMatrix * a_position; \n\
position.x += translate.x; \n\
position.y += translate.y; \n\
gl_Position = CC_PMatrix * position; \n\
\n\
vec3 normal = (CC_MVMatrix * vec4(a_normal, 1.0)).xyz; \n\
vec3 n_normal = normalize(normal); \n\
vec3 n_lightDirection = normalize(lightDirection); \n\
float cosNL = max(dot(n_normal, n_lightDirection), 0.0); \n\
vec4 diffuse = vec4(lightDiffuse * materialDiffuse, 1.0) * cosNL; \n\
\n\
vec4 ambient = vec4(lightAmbient * materialAmbient, 1.0); \n\
\n\
vec3 n_eyePosition = normalize(eyePosition - position.xyz); \n\
vec3 reflection = 2 * max(dot(n_normal, n_lightDirection), 0.0) * n_normal - n_lightDirection; \n\
vec4 specular = vec4(lightSpecular * materialSpecular * pow(max(dot(n_eyePosition, reflection), 0.0), shininess), 1.0); \n\
\n\
v_color = a_color * (diffuse + ambient + specular); \n\
v_texCoord = a_texCoord; \n\
} \n\
";
const char* ccPositionTextureColorForLight1_f = " \
#ifdef GL_ES \n\
precision lowp float; \n\
#endif \n\
\n\
varying vec4 v_color; \n\
varying vec2 v_texCoord; \n\
\n\
void main() { \n\
vec4 color = v_color * texture2D(CC_Texture0, v_texCoord); \n\
color.a = 1.0; \n\
\n\
gl_FragColor = color; \n\
} \n\
";

效果:

// 在片段着色器中处理光照计算(环境光+漫反射+镜面反射)
const char* ccPositionTextureColorForLight2_v = " \
uniform vec2 translate; \n\
\n\
attribute vec4 a_position; \n\
attribute vec4 a_color; \n\
attribute vec2 a_texCoord; \n\
attribute vec3 a_normal; \n\
\n\
#ifdef GL_ES \n\
varying mediump vec4 v_position; \n\
varying mediump vec4 v_color; \n\
varying mediump vec2 v_texCoord; \n\
varying mediump vec3 v_normal; \n\
#else \n\
varying vec4 v_position; \n\
varying vec4 v_color; \n\
varying vec2 v_texCoord; \n\
varying vec3 v_normal; \n\
#endif \n\
\n\
void main() { \n\
vec4 position = CC_MVMatrix * a_position; \n\
position.x += translate.x; \n\
position.y += translate.y; \n\
gl_Position = CC_PMatrix * position; \n\
\n\
v_position = position; \n\
v_color = a_color; \n\
v_texCoord = a_texCoord; \n\
v_normal = (CC_MVMatrix * vec4(a_normal, 1.0)).xyz; \n\
} \n\
";
const char* ccPositionTextureColorForLight2_f = " \
const vec3 lightAmbient = vec3(1.0, 1.0, 1.0); \n\
const vec3 lightDiffuse = vec3(1.0, 1.0, 1.0); \n\
const vec3 lightSpecular = vec3(1.0, 1.0, 1.0); \n\
const vec3 materialAmbient = vec3(1.0, 1.0, 1.0); \n\
const vec3 materialDiffuse = vec3(1.0, 1.0, 1.0); \n\
const vec3 materialSpecular = vec3(1.0, 1.0, 1.0); \n\
const int shininess = 8; // 0-128 \n\
\n\
const vec3 eyePosition = vec3(512.0, 128.0, 442.0); \n\
const vec3 lightDirection = vec3(1.0, 1.0, 1.0); \n\
\n\
#ifdef GL_ES \n\
precision lowp float; \n\
#endif \n\
\n\
varying vec4 v_position; \n\
varying vec4 v_color; \n\
varying vec2 v_texCoord; \n\
varying vec3 v_normal; \n\
\n\
void main() { \n\
vec4 color = v_color * texture2D(CC_Texture0, v_texCoord); \n\
color.a = 1.0; \n\
\n\
vec3 normal = v_normal; \n\
vec3 n_normal = normalize(normal); \n\
vec3 n_lightDirection = normalize(lightDirection); \n\
float cosNL = max(dot(n_normal, n_lightDirection), 0.0); \n\
vec4 diffuse = vec4(lightDiffuse * materialDiffuse, 1.0) * cosNL; \n\
\n\
vec4 ambient = vec4(lightAmbient * materialAmbient, 1.0); \n\
\n\
vec3 n_eyeDirection = normalize(eyePosition - v_position.xyz); \n\
vec3 reflection = 2 * max(dot(n_normal, n_lightDirection), 0.0) * n_normal - n_lightDirection; \n\
vec4 specular = vec4(lightSpecular * materialSpecular * pow(max(dot(n_eyeDirection, reflection), 0.0), shininess), 1.0); \n\
//vec3 reflection = normalize(n_eyeDirection + n_lightDirection); \n\
//vec4 specular = vec4(lightSpecular * materialSpecular * pow(max(dot(n_normal, reflection), 0.0), shininess), 1.0); \n\
\n\
gl_FragColor = color * (ambient + diffuse + specular); \n\
} \n\
";

效果:

// 在片段着色器中处理光照计算(环境光+漫反射+镜面反射+衰减系数)
const char* ccPositionTextureColorForLight3_v = " \
uniform vec2 translate; \n\
\n\
attribute vec4 a_position; \n\
attribute vec4 a_color; \n\
attribute vec2 a_texCoord; \n\
attribute vec3 a_normal; \n\
\n\
#ifdef GL_ES \n\
varying mediump vec4 v_position; \n\
varying mediump vec4 v_color; \n\
varying mediump vec2 v_texCoord; \n\
varying mediump vec3 v_normal; \n\
varying float v_distance; \n\
#else \n\
varying vec4 v_position; \n\
varying vec4 v_color; \n\
varying vec2 v_texCoord; \n\
varying vec3 v_normal; \n\
varying float v_distance; \n\
#endif \n\
\n\
void main() { \n\
vec4 position = CC_MVMatrix * a_position; \n\
position.x += translate.x; \n\
position.y += translate.y; \n\
gl_Position = CC_PMatrix * position; \n\
\n\
v_position = position; \n\
v_color = a_color; \n\
v_texCoord = a_texCoord; \n\
v_normal = (CC_MVMatrix * vec4(a_normal, 1.0)).xyz; \n\
} \n\
";
const char* ccPositionTextureColorForLight3_f = " \
const vec3 lightAmbient = vec3(1.0, 1.0, 1.0); \n\
const vec3 lightDiffuse = vec3(1.0, 1.0, 1.0); \n\
const vec3 lightSpecular = vec3(1.0, 1.0, 1.0); \n\
const float lightAttenuationConst = 1.0; \n\
const float lightAttenuationLinear = 0.0002; \n\
const float lightAttenuationQuadratic = 0.000001; \n\
const vec3 materialAmbient = vec3(1.0, 1.0, 1.0); \n\
const vec3 materialDiffuse = vec3(1.0, 1.0, 1.0); \n\
const vec3 materialSpecular = vec3(1.0, 1.0, 1.0); \n\
const int shininess = 8; // 0-128 \n\
\n\
const vec3 lightPosition = vec3(1024.0, 512.0, 100.0); \n\
const vec3 eyePosition = vec3(512.0, 256.0, 442.0); \n\
\n\
#ifdef GL_ES \n\
precision lowp float; \n\
#endif \n\
\n\
varying vec4 v_position; \n\
varying vec4 v_color; \n\
varying vec2 v_texCoord; \n\
varying vec3 v_normal; \n\
varying float v_distance; \n\
\n\
void main() { \n\
vec4 color = v_color * texture2D(CC_Texture0, v_texCoord); \n\
color.a = 1.0; \n\
\n\
vec3 lightDirection = lightPosition - v_position.xyz; \n\
float distance = length(lightDirection); \n\
float atten = 1.0 / (lightAttenuationConst + lightAttenuationLinear * distance + lightAttenuationQuadratic * distance * distance); \n\
\n\
vec3 normal = v_normal; \n\
vec3 n_normal = normalize(normal); \n\
vec3 n_lightDirection = normalize(lightDirection); \n\
float cosNL = max(dot(n_normal, n_lightDirection), 0.0); \n\
vec4 diffuse = vec4(lightDiffuse * materialDiffuse, 1.0) * cosNL; \n\
\n\
vec4 ambient = vec4(lightAmbient * materialAmbient, 1.0); \n\
\n\
vec3 n_eyeDirection = normalize(eyePosition - v_position.xyz); \n\
vec3 reflection = 2 * max(dot(n_normal, n_lightDirection), 0.0) * n_normal - n_lightDirection; \n\
vec4 specular = vec4(lightSpecular * materialSpecular * pow(max(dot(n_eyeDirection, reflection), 0.0), shininess), 1.0); \n\
//vec3 reflection = normalize(n_eyeDirection + n_lightDirection); \n\
//vec4 specular = vec4(lightSpecular * materialSpecular * pow(max(dot(n_normal, reflection), 0.0), shininess), 1.0); \n\
\n\
float colorA = color.a; \n\
gl_FragColor = vec4(color.rgb * (ambient + diffuse + specular).xyz * atten, colorA); \n\
} \n\
";

效果:

// 在片段着色器中处理光照计算(环境光+漫反射+镜面反射+聚光灯+衰减系数)
const char* ccPositionTextureColorForLight4_v = " \
uniform vec2 translate; \n\
const vec3 spotlightPosition = vec3(512.0, 256.0, 120.0); \n\
const vec3 lightPosition = vec3(512.0, 256.0, 442.0); \n\
const vec3 eyePosition = vec3(512.0, 256.0, 442.0); \n\
\n\
attribute vec4 a_position; \n\
attribute vec4 a_color; \n\
attribute vec2 a_texCoord; \n\
attribute vec3 a_normal; \n\
\n\
#ifdef GL_ES \n\
varying mediump vec4 v_position; \n\
varying mediump vec4 v_color; \n\
varying mediump vec2 v_texCoord; \n\
varying mediump vec3 v_normal; \n\
varying mediump float v_distance; \n\
varying mediump vec3 v_lightDirection; \n\
varying mediump vec3 v_spotlightDirection; \n\
#else \n\
varying vec4 v_position; \n\
varying vec4 v_color; \n\
varying vec2 v_texCoord; \n\
varying vec3 v_normal; \n\
varying float v_distance; \n\
varying vec3 v_lightDirection; \n\
varying vec3 v_spotlightDirection; \n\
#endif \n\
\n\
void main() { \n\
vec4 position = CC_MVMatrix * a_position; \n\
position.x += translate.x; \n\
position.y += translate.y; \n\
gl_Position = CC_PMatrix * position; \n\
\n\
v_position = position; \n\
v_color = a_color; \n\
v_texCoord = a_texCoord; \n\
v_normal = (CC_MVMatrix * vec4(a_normal, 1.0)).xyz; \n\
\n\
vec3 lightDirection = lightPosition - position.xyz; \n\
vec3 spotlightDirection = spotlightPosition - position.xyz; \n\
v_distance = length(lightDirection); \n\
v_lightDirection = normalize(lightDirection); \n\
v_spotlightDirection = normalize(spotlightDirection); \n\
} \n\
";
const char* ccPositionTextureColorForLight4_f = " \
const vec3 lightAmbient = vec3(1.0, 1.0, 1.0); \n\
const vec3 lightDiffuse = vec3(1.0, 1.0, 1.0); \n\
const vec3 lightSpecular = vec3(1.0, 1.0, 1.0); \n\
const float lightAttenuationConst = 1.0; \n\
const float lightAttenuationLinear = 0.001; \n\
const float lightAttenuationQuadratic = 0.000001; \n\
const vec3 materialAmbient = vec3(1.0, 1.0, 1.0); \n\
const vec3 materialDiffuse = vec3(1.0, 1.0, 1.0); \n\
const vec3 materialSpecular = vec3(1.0, 1.0, 1.0); \n\
const int shininess = 8; // 0-128 \n\
const int lightSpotExponent = 1; \n\
const float lightSpotCosCutoff = cos(45.0 / 180.0 * 3.1415926); \n\
\n\
const vec3 lightPosition = vec3(512.0, 256.0, 442.0); \n\
const vec3 eyePosition = vec3(512.0, 256.0, 442.0); \n\
\n\
#ifdef GL_ES \n\
precision lowp float; \n\
#endif \n\
\n\
varying vec4 v_position; \n\
varying vec4 v_color; \n\
varying vec2 v_texCoord; \n\
varying vec3 v_normal; \n\
varying float v_distance; \n\
varying vec3 v_lightDirection; \n\
varying vec3 v_spotlightDirection; \n\
\n\
void main() { \n\
vec4 color = v_color * texture2D(CC_Texture0, v_texCoord); \n\
color.a = 1.0; \n\
\n\
// vec3 lightDirection = lightPosition - v_position.xyz; \n\
vec3 n_spotDirection = normalize(lightPosition - vec3(512.0, 256.0, 0.0)); \n\
vec3 n_spotlightDirection = normalize(v_spotlightDirection); \n\
// float distance = length(lightDirection); \n\
float atten = 1.0 / (lightAttenuationConst + lightAttenuationLinear * v_distance + lightAttenuationQuadratic * v_distance * v_distance); \n\
\n\
vec3 normal = v_normal; \n\
vec3 n_normal = normalize(normal); \n\
vec3 n_lightDirection = normalize(v_lightDirection); \n\
float cosNL = max(dot(n_normal, n_lightDirection), 0.0); \n\
vec4 diffuse = vec4(lightDiffuse * materialDiffuse, 1.0) * cosNL; \n\
\n\
vec4 ambient = vec4(lightAmbient * materialAmbient, 1.0); \n\
\n\
vec3 n_eyeDirection = normalize(eyePosition - v_position.xyz); \n\
vec3 reflection = 2 * max(dot(n_normal, n_lightDirection), 0.0) * n_normal - n_lightDirection; \n\
vec4 specular = vec4(lightSpecular * materialSpecular * pow(max(dot(n_eyeDirection, reflection), 0.0), shininess), 1.0); \n\
//vec3 reflection = normalize(n_eyeDirection + n_lightDirection); \n\
//vec4 specular = vec4(lightSpecular * materialSpecular * pow(max(dot(n_normal, reflection), 0.0), shininess), 1.0); \n\
\n\
float cosSpot = dot(-1 * n_spotDirection, -1 * n_spotlightDirection); \n\
if (cosSpot > lightSpotCosCutoff) \n\
{ \n\
float spotEffect = pow(cosSpot, lightSpotExponent); \n\
atten += spotEffect; \n\
} \n\
float colorA = color.a; \n\
gl_FragColor = vec4(color.rgb * (ambient + diffuse + specular).xyz * atten, colorA); \n\
} \n\
";

效果:

GLSL 中的光照计算的更多相关文章

  1. GLSL逐像素光照 【转】

    转载:http://blog.csdn.net/hgl868/article/details/7872414 逐像素的方向光(Directional Light per Pixel) 这一节将把前面的 ...

  2. GLSL逐顶点光照[转]

    转载:http://blog.csdn.net/hgl868/article/details/7872350 引言 在OpenGL中有三种类型的光:方向光(directional).点光(point) ...

  3. 【Unity Shaders】Shader中的光照

    写在前面 自己写过Vertex & Fragment Shader的童鞋,大概都会对Unity的光照痛恨不已.当然,我相信这是因为我们写得少...不过这也是由于官方文档对这方面介绍很少的缘故, ...

  4. 关于Unity中的光照(七)

    全局光照 GI 这里所说的反射就是,一个红色的物体,当太阳照射它的时候,它周围的物体也会变得有点红. 1:Realtime每帧都会计算光照,实时光照是不会反射的,所以它的光影显得单调;2:Baked ...

  5. 关于Unity中的光照(五)

    Mobile Diffuse Unity自带的一种shader,用的比较多,性能还可以.我们默认创建的unit shader基本和它一致,但是没有参与光照计算,看起来和Mobile Diffuse有区 ...

  6. OpenGL中的光照技术(翻译)

    Lighting:https://www.evl.uic.edu/julian/cs488/2005-11-03/index.html 光照 OpenGL中的光照(Linghting)是很重要的,为什 ...

  7. Deferred Shading,延迟渲染(提高渲染效率,减少多余光照计算)【转】

    Deferred Shading,看过<Gems2> 的应该都了解了.最近很火的星际2就是使用了Deferred Shading. 原帖位置:   http://blog.csdn.net ...

  8. Opengl场景中加光照包含几个步骤

    http://zuoye.baidu.com/question/44e2a82d7ad5c0e1d33ddb9a40e0bf86.html  Opengl场景中加光照包含几个步骤,各个步骤实现用的函数 ...

  9. Shader 中的颜色计算

    下面介绍 Shader 中 gl_FragColor 的计算与转换: 一.颜色计算 1. 加 这里要讲讲三原色和三基色:三原色一般指的是红.绿.蓝三种,简称 RGB,这是加色系.就是光源只含有特定的波 ...

随机推荐

  1. 深度优化LNMP

    优化前准备工作 Centos准备及配置 准备安装包及软件:http://pan.baidu.com/s/1chHQF  下载解压到U盘即可安装http://pan.baidu.com/s/15TUWf ...

  2. ADF_ADF Faces系列2_使用JSF开发基于Ajax的用户界面:ADF Faces富客户端组件简介(Part2)

    2013-05-01 Created By BaoXinjian

  3. bzoj2005 能量采集 gcd 容斥

    ans = sigma_x(sigma_y(gcd(x,y) * 2 - 1)),1<=x<=n,1<=y<=m 枚举x,y,O(nmlogn),超时 换个角度,枚举d = g ...

  4. [实变函数]5.3 非负可测函数的 Lebesgue 积分

    本节中, 设 $f,g,f_i$ 是可测集 $E$ 上的非负可测函数, $A,B$ 是 $E$ 的可测子集.       1 定义: (1) $f$ 在 $E$ 上的 Lebesgue 积分      ...

  5. JAVA 边界布局管理器

    //边界布局管理器 import java.awt.*; import javax.swing.*; public class Jiemian1 extends JFrame{ //定义组件 JBut ...

  6. JAVA关系运算符

    常用的关系运算符: 编号 关系运算符 说明 1 > 大于 2 < 小于 3 == 等于 4 != 不等于 5 >= 大于或等于 6 <= 小于或等于 用关系运算符判断后返回的值 ...

  7. 保持UIImagePickerController后导航栏风格统一

    1. UIImagePickerController 状态栏始终保持某一种风格. -(void)navigationController:(UINavigationController *)navig ...

  8. mysql 按年度、季度、月度、周、日SQL统计查询

    创建Table CREATE TABLE `test` ( `cdate` datetime DEFAULT NULL, `id` ) DEFAULT NULL, `name` ) DEFAULT N ...

  9. [Java] java中的接口定义

    在Java的通常规范中,对数据成员的修改要通过接口提供的方法进行(如下面示例中接口中的void learnMath(int hours)和void learnEnglish(int hours)),这 ...

  10. 使用 Heka 导入自定义的nginx日志到Elasticsearch

    重置Heka执行进度 heka的进度配置文件存在配置项 base_dir 设置的目录,只需要删除这个文件夹下面的内容,就可以完全重置heka的进度. base_dir 配置项默认是在下面目录: '/v ...