GRAPHICS PROGRAMMING Basic of computer graphics with OpenGL Handful graphics function OpenGL : by silicon graphics PHIGS : Programmer’s Hierarchical Graphics System GKS : Graphic Kernel System JAVA-3D By Sun micro-system DirectX: Microsoft corp. WHY 3D Former graphic model Pen plotter model Useful for drawing 2D large diagram moveto(0,0); API ex. LOGO, GKS, and PostScript lineto(1,0); Unsuitable for 3D model : need sophisticated lineto(0,0); math to user lineto(1,1); lineto(0,1); moveto(0,1); lineto(0.5,1.866); lineto(1.5,1.866); lineto(1.5, 0.866); lineto(1,0); moveto(1,1); lineto(1.5, 1.866); Coordinate Systems CG system is unable define exactly unit like cm, inch etc CG is a device independent system Current coordinate is user coordinate = world coordinate It should be match with CRT coordinate system (Raster coordinate) Graphic function properties 7 groups of function Primitive: What is object ? low level objects or atomic entities, ex. point, polygon etc, Attribute How the appear: fill, bold character Viewing How we saw the image Transformation Transform of object: rotate, move Input Deal with the devices: keyboard, mouse etc. Control function Multiwindow, multiprocessing environment handling. Inquiry function Information providing for different API Pipeline and State Machine Entire graphic system thinking as a state machine There are 2 types of Graphic functions thing that define primitives thing that changes the state The OpenGL Interface Begin with “gl” Stored in library and referred to as GL There are Graphics Utility Library (GLU) GLU Toolkit (GLUT) GLX or WGL : glue for GL to OS Defined in standard header folder “GL” filename “glut.h” Primitives and Attributes API should contain small set of primitives that every hardware can be supported Ex. Line, polygons, text Variety of primitive such as circle, curves, surface and solids able to build sophisticated object but few hardware supported OpenGL takes an intermediate Support 2 classes of primitives Geometric primitives : pass through a geometric pipeline Raster primitives: pass through pixel pipeline Geometric Able to manipulated Raster Lack of geometric properties Let’s have a look at 2D Modeling Special case of 3D Suppose z=0, every point refer to (x, y,0) Generally object created from set points In graphics system , the word “vertex” more preferred that “point” OpenGL function form glVertex*(); where *: nt or ntv , 2 or 3 characters form n : number of dimension ( 2, 3 or 4) t : data type (ingeter, float, double, v for pointer) Ex. glVertex2i(); /* vertex for 2D integer type*/ The data type may change to GL type instead of C Ex. GLfloat = float in C Note: All of them have already defined in header fine <GL\glut.h> OpenGL Object form Defined object in glBeginglEnd loop 2 kinds of primitives that is used to defined object No interior, eg. points, line glBegin(type); glVertex*(…); . . . glEnd(); C command for defining object have surface, eg. polygon Difference type of object form Polygon Basics Close object that has interior Able to use as curve surface Number of generated polygons per time is used as graphic performance Display either only edges or fill Correct properties should be simple, convex, and flat 3D polygon is unnecessarily flat Filled objects polygons displaying simple polygon nonsimple polygon convex property Convex object properties 3D convex object: 3 vertices are not collinear Safe for rendering if use triangle Hardware and software often support Types of Polygon GL_POLYGONS Edges are perform line loop and close Edges has no with define either fill or edges using glPolygonMode If both, draw twice Special types polygon Triangles and Quadrilaterals (GL_TRIANGLES, GL_QUADS) Strips and Fans (GL_TRIANGLE_STRIP, GL_QUAD_STRIP, GL_TRIANGLE_FAN) Sample object: Generating a Sphere assign to be polygons and used GL_QUAD_STRIP Use longitude and latitude schemes for the middle body For pole uses GL_TRIANGLE_FAN float C= PI/180.0; //degrees to radians, M_PI = 3.14159… for (float phi = -80.0; phi <= 80.0; phi += 20.0) { glBegin(GL_QUAD_STRIP); for (float theta = -180.0; theta <= 180.0; theta += 20.0) { float x=sin(c*theta)*cos(c*phi); float y=cos(c*theta)*cos(c*phi); float z=sin(c*phi); glVertex3d(x,y,z); x=sin(c*theta)*cos(c*(phi+20.0)); y=cos(c*theta)*cos(c*(phi+20.0)); z=sin(c*(phi+20.0)); glVertex3d(x,y,z); } glEnd(); } x=y=0; // North pole modeling z = 1; glBegin(GL_TRIANGLE_FAN); glVertex3d(x,y,z); c=M_Pi/180.0; z=sin(c*80.0); for (theta=-180.0; theta<=180.0;theta+=20.0) { x=sin(c*theta)*cos(c*80.0); y=cos(c*theta)*cos(c*80.0); glVertex3d(x,y,z); } glEnd(); x=y=0; // South pole modeling z=-1: glBegin(GL_TRIANGLE_FAN); glVertex3d(x,y,z); z = -sin(c*80.0); for(theta = -180.0; theta <= 180.0; theta=20.0) { x=sin(c*theta)*cos(c*80.0); y=cos(c*theta)*cos(c*80.0); glVertex3d(x,y,z); } glEnd(); Today topic fonts attributes color system drawing a graphic with OpenGL Text 2 types of text Stroke Text Constructed via using graphic primitives Able to transform like other primitives Raster Text Character are defined as rectangle of bits block Stroke text Consume a lot of memories Postscript as an example Raster text Rapidly be placed in buffer by using bit-block-transfer (bitblt) operation Operate only character sizing Often store in ROM (hardware) Portability is limited by particular font GLUT provide 8x8 pixels function glutBitmapCharacter(GLUT_BITMAP_8_BY_13, C) C: ASCII character number Character is placed in the present position of screen Curved Objects Create by using 2 approach Use the primitive except points n side polygon instead of circle Approximate sphere with polyhedron Curved surface by a mesh of convex polygon Use mathematical definition Quadric surfaces and parametric polynomial curved and surfaces example: Define sphere by center and a point on surface Cubic Polynomial is defined by 4 points OpenGL able to do both Attributes About how primitive display Line : display color, type of line (dash, solid) Concern with immediate mode: display as soon as they are defined Color Most interesting of perception and computer graphics Base on three color theory If using additive color model - c = T1R+T2G+T3B C: color that we trying to match T1, T2, T3: strength of intensity, the tristimulus value Human Visual System Our visual system do a continuous perception Depends on 3 types of cone cell Visually indistinguishable if they have the same tristimulus value CRT is an example of additive color system Ai Si ( )C ( )d Ai: brain perception value Si: cone cell sensitivity Viewing a point as a color solid cube Subtractive color model The complementary of additive color model Start with white surface If white light hit the surface, color will be absorb except the object color which are reflect Ex. painting and printing Complementary color: cyan, magenta, yellow additive color model subtractive color model RGB-color model Use separate buffer for each color Each pixel has 3 bytes (24 bits) for each color 16 Million shade of color OpenGL function glColor3f(r, g, b); ex. Red glColor3f(1.0, 0.0, 0.0); RGBA, the 4 color model A: Alpha channel Store in frame buffer like RGB For creating effect ex. fog, combining images. OpenGL treat as opacity or transparency Ex. OpenGL command for 4 color model glClearColor(1.0, 1.0, 1.0, 1.0); White color and opaque Indexed Color Difficult to support in hardware Higher memory requirements but now memory is cheaper Use color tray of artist as principle Infinite color can be produced from different quantity of primary colors OpenGL indexed color function glIndex(element); Select color out of table glutSetcolor(int color, GLfloat red, GLfloat blue, GLfloat green); Set color entry to map the color table Color Attributes For RGB mode glClearColor(1.0, 1.0, 1.0); /* clear to white */ glColor3f(1.0, 0.0, 0.0); /* setting point to red */ glPointSize(2.0); /* 2 pixel wide */ Note: If 2 display differ in pixel size, rendered images may appear slightly different Viewing Method for objects appear on screen Use synthetic camera concept Fix lens and fix location Picture would be distort like real world If we need to take an elephant picture, camera should far enough to take all information 2D Viewing Base on the 2D rectangular area Know as viewing rectangle or clipping rectangle Be a special case of 3D viewing ex. plane at z=0 Default in 2x2x2 volume, origin in the center and bottom-left corner is at (-1.0, -1.0) Orthographic View 2D view the orthographic projection of 3D Function void glOrtho(GLdouble left, GLdouble right, GLdouble bottom, GLdouble top, GLdouble near, GLdouble far); // near, far: distance which are measured from camera /* orthographic projection from 3D */ void gluOrtho2D(GLdouble left, GLdouble right, GLdouble bottom, GLdouble top); /* 2D equivalent to glOrtho but near and far set to -1.0, 1.0 */ Unlike camera, it is able to view behind object Matrix Modes Between graphic pipeline state, any transformation 2 important matrices: model-view Projection glMatrixMode(GL_PROJECTION); glLoadIdentity(); gluOrtho2d(0.0, 500.0, 0.0, 500.0); glMatrixMode(GL_MODELVIEW); Control function Concern about software environment between software and platform Different platform will have different interfacing GLUT also provide the utility : see further Windows interfacing Window : a rectangular area of our display, max = CRT screen Window default origin: (0,0) at lower-left corner like CRT but mouse at top-left OpenGL function (GLUT function) for window glutInit (int *argcp, char **argv); glutCreateWindow(char *title); /* given the window title */ Change the display setup glutInitDisplayMode(GLUT_RGB| GLUT_DEPTH | GLUT_DOUBLE); GLUT_RGB: define RGB color mode GLUT_DEPTH: a depth buffer for hidden-surface removal GLUT_DOUBLE: number of buffer Double/Single default: RGB color, no hidden surface removal, single buffering glutInitWindowSize(480, 640); glutInitWindowPosition(0,0) Aspect ratio Ratio of rectangle’s width to its height If glOrtho and glutInitWindowSize are not specified the same size, object are distort. View port A rectangular area of the display window Setting a view port void glViewport(GLint x, GLint y, GLsizei w, GLsizei h); The function: main, display and myinit glutMainLoop(); /* begin an event-processing loop, let the window waiting for kill process */ void glutDisplayFunc(void *(func)(void)); /* call to the redisplay function name func */ #include <GL/glut.h> void main(int argc, char **argv){ glutInit(&argc, argv); glutInitDisplayMode(GLUT_SINGLE | GLUT_RGB ); glutInitWindowSize(500, 500); glutInitWindowPosition(0, 0); glutCreateWindow("Simple OpenGL example"); glutDisplayFunc(display); myinit(); glutMainLoop(); } Program template Program structure consisting myinit : setup user options to state variables dealing with viewing and attributes-parameters Example program: Sierspinski Gasket Proceeding of Sierspinski 1. 2. 3. 4. 5. 6. Pick an random initial point in triangle Select vertex Finding the halfway point between initial point and random vertex Mark and display new point Replace the initial point with this new point Return to step 2 Pseudo code main() { Initialize_the_system(); for(some_number_of_points) { pt = generate_a_point(); display_the_point(pt); } cleanup(); Sierpinski gasket } Using array with OpenGL // For 3D vertex, 2D is a special case GLfloat vertex[3]; /* define array */ // Then we can use glVertex3fv(vertex); /* pass by reference */ // Defining geometric object in Begin and End fn. statement glBegin(GL_LINES); glVertex2f(x1, y1); glVertex2f(x2, y2); glEnd(); The same data able to define another object // define a pair of points glBegin(GL_POINTS); glVertex2f(x1, y1); glVertex2f(x2, y2); geEnd(); Using a 2 element array to carry point a // By defining new data type with 2 element array typedef GLfloat point2[2]; // point2[0] carry x data // point2[1] carry y data when use point2 vertices[3] ; // the members are vertices[0][0], vertices[1][0], vertices[2][0] // carry x value vertices[0][1], vertices[1][1], vertices[2][1] // carry y value point2 vertices[3] = {{0.0, 0.0}, {250.0, 500.0}, {500,0}} Thing need to do Coloring Locate the image Define size Window creating Image clipping Image duration 5000 random point 2D Sierspinski Triangular gasket There is no point in the middle triangle The same observation can be applied to the other 3 triangles and so on Another method to fill the area is use triangle polygon instead of point Strategy Start with a triangle which subdivide the area to 4 triangles Remove the middle one Repeat to other triangles until the size of the removing triangle is small enough. Let say 1 pixel This is the recursive program See program typedef float point2[2]; /* initial triangle */ point2 v[]={{-1.0, -0.58}, {1.0, -0.58}, {0.0, 1.15}}; void triangle( point2 a, point2 b, point2 c) { /* display one triangle */ glBegin(GL_TRIANGLES); glVertex2fv(a); glVertex2fv(b); glVertex2fv(c); glEnd(); } void divide_triangle(point2 a, point2 b, point2 c, int m) { /* triangle subdivision using vertex numbers */ point2 v0, v1, v2; int j; if (m>0) { for(j=0; j<2; j++) v0[j]=(a[j]+b[j])/2; for(j=0; j<2; j++) v1[j]=(a[j]+c[j])/2; for(j=0; j<2; j++) v2[j]=(b[j]+c[j])/2; divide_triangle(a, v0, v1, m-1); divide_triangle(c, v1, v2, m-1); divide_triangle(b, v2, v0, m-1); } else /* draw triangle at end of recursion */ triangle(a,b,c); } void display(void) { glClear(GL_COLOR_BUFFER_BIT); divide_triangle(v[0], v[1], v[2], n); glFlush(); } with 4 level recursion with 5 level recursion 3D Sierspinski gasket Begin with tetrahedron instead triangle Use 3D point point v[]={ { 0.0, 0.0, 1.0}, { 0.0, 0.942809, -0.33333}, {-0.816497, -0.471405, -0.333333}, { 0.816497, -0.471405, -0.333333} }; The hidden surface removal Problem may happen if there is no relation between surface Close opaque object should mask the far object The part of far object which overlap with close object should remove Z-buffer algorithm is a method to manipulate. Exercises Write a part of C program to define a unit circle object at position (1,1) using OpenGL command Hint: you may use primitive such as TRIANGLE_FANS, or others