I'm starting study opengl, and im tring to make a 3d chess like, but i cant figureout, how i can know where i have clicked in the "table" to make the proper animations, any advice ?
This is called "3D picking". You have to translate screen coordinates into world coordinates. From there, do a ray/collision object (bounding box?) intersection test. If they intersect, that's where the user clicked.
You'll have to do a little bit more than this to solve the depth-order problem, like finding the first time to intersection of each object, then selecting the one with the lowest (positive) time.
If you google for "3D picking" you might find what you are looking for.
Here is a tutorial:
http://nehe.gamedev.net/data/lessons/lesson.asp?lesson=32
Note that this is not specific to any shape of bounding object, be it a bounding box, a polygon, a curve, etc. You just have to figure out the math for the intersection test for each type of object you want to support.
Edit:
I didn't read that tutorial before I linked it, I just figured NEHE is where all the cool kids learn OpenGL (admittedly ten years ago...).
Here is something from the OpenGL FAQ about picking:
http://www.opengl.org/resources/faq/technical/selection.htm
waldecir, look for a raypick function. It's the name for sending a ray from the scene's camera center through the pixel you clicked on (actually, through that pixel's translated position on the camera's plane representing the "glass surface of the screen" in the 3D world) and return the frontmost polygon the ray hits together with some information. Usually coordinates within the polygon's surface axes, e.g. UV or texture coordinates. By checking the coordinates, you can determine which square the user clicked on.
Rays can be sent from any position and in any direction, so likely you'd have to get the camera position and its plane center, but the documentation should be able to help you there.
Related
I want to make a terrain where the ending point is also the starting point. So, like on the earth you could just go on walking straight and you would reach the point where you started again after some time.
Thanks for your help!
Unity's Terrain system can only create square regions of terrain. So this can't be done as such.
However, you can approximate it, and I'll tell you how I've done it in my project to some success.
Figure out how much terrain you need to cover the "globe", we'll say it takes NxN chunks of terrain we'll call a "tile".
What you do next is you make 9 of those NxN tiles, and arrange them in a 3x3 grid. Put the camera in the center tile of the grid, and whenever the camera leaves that tile, determine where it is on the tile it is on, then change its position to the corresponding position on the center tile.
This will give you a "toroidal" world. I found this was the easiest solution to get the player to see things on the other "corner" of the world map, and then cross into it without graphical issues.
If you have other objects residing on the world, that presents some additional challenges. One thing you can start with is duplicating them 9x and start them at the same relative position of each tile. If they only interact with the player, that should be fine, just whenever the player interacts with 1, the other 8 do whatever that 1 does.
If the other residents of the globe have to interact with each other, you'll need a way to figure out how to make all 9 copies of everything behave consistently, but that's too broad of a question to address here.
After using Unity for over a year now for creating 3D games I'd like to do my first pseudo 3D or 2.5D game (or whatever it's called). What I mean is games like Clash of Clans or Boom Beach where objects are really 2D images rendered to give 3D feeling. Because I don't even know the correct word for that type of games it's really hard to get started. What I can think of is giving Camera rotation of (45, -45, 0) so that it looks down at the ground from the upper right corner and then creating empty gameObjects with Sprite Renderers and setting their rotation to (22.5, -45, 0) so that they face the camera. I don't know if that's how it's really done so could somebody give me a link to some neat tutorial or something because I'm stuck.
The keyword you are looking for is isometric projection.
You can use Unity to create that effect with 3D graphics. You need to set the camera rotation to (45, -45, 0) and set the Camera Projection to Orthographic.
If you want to only use 2D graphics then the graphics are made in such a way that they are drawn in an angle that it looks like you are looking at them in 45 degrees.
If you search for isometric unity then you should find a ton of tutorials for Unity.
How can I override the physical marker in unity to just make the physical marker invisible in camera while detecting any object.
Like in this video:
https://m.youtube.com/watch?v=R_F1LvK5gCk
And many other videos.
This maybe isn't the correct answer but, I'll put my ideas here because sadly it doesn't fit in the comment section.
I just watch the video and some of the reels they show as his works.
First, yes, in deed they make the marker a sort of invisible marker, but if you look closely, like a 0.25 speed and pushing space bar very fast to look it in slowmo, you can see a kind of "artifact" in between the girl's fingers that makes me think there's no invisible marker but a texture that are covering the marker. Maybe a cylinder that gets his texture from the video camera input.
Now, how I do that?
There are several ways to get the pixels from the webcamera, even Unity has one function. The trouble is, I dont want all the pixels but a little tiny part of the camera render, specifically the one's around my marker.
In my experience, and in his examples, they are using OpenCV, another unity plugin, so they can track anything, from faces, hands, or markers, so I can't be quite sure they are using Vuforia alone or in combination.
My idea is, with OpenCV you can catch your marker and his contour, then ask for the pixels outside the contour of your marker, those pixels will be the skin tone of the person and latter apply them as a texture over some plane or 3d model that can cover your marker. You can have the pixels at right side and left side of the marker and use a average function betwen them so it could look nice or if you like the adventures you can try to use some kind of digital images processing method to get bether results.
I'm not sure if you can have the pixels arround a marker just using just Vuforia. Honestly I've never try it before.
Well, that's my idea.
If you can get it better I'll like to hear about it.
I am working in Unity3D and I was wondering how I would rotate a cube based on the angle between the cube and the mouse position. With that I don't mean the mouse position in world space but in pixels.
Here are some pages that'll lead you to your answer:
Input.mousePosition This also includes an example of how to turn screen coordinates into a ray in world coordinates. If you know how far away from the camera you want your point, check out ScreenToWorldPoint for a point instead of a ray.
transform.Rotate To perform a rotation.
The rest of your question is kinda vague--rotating "based on" the angle between cube and mouse position could mean a lot of things. You shouldn't need much more information than this though!
I have created a 3d environment full of 3D cubes, does anyone have any idea how you would detect a touch on one of these Cubes. I thinking if I could get the cubes screen position (coords start from bottom left) then it would be pretty easy
UPDATE:
I added the function -(CGPoint)getScreenCoorOfPoint:(IMPoint3D)_point3D which seems to give me my items position in the world but the bit I am now stuck on is:
I have objects that have a position
I have my position in the world (gluLookAt eye[0], eye[1], eye[2])
and then I have where I tapped on the screen
How do I join all this up, its the last thing in my way to archiving greatness!!!!
Look up OpenGL picking on Google. There are two main methods to accomplish this, I recommend you use the second one described at OpenGL.org as it does not involve rendering anything offscreen:
[…] involves shooting a pick ray through the mouse location and testing for intersections with the currently displayed objects. OpenGL doesn't test for ray intersections, but you'll need to interact with OpenGL to generate the pick ray.
Also see this question for some discussion on the matter:
Screen-to-World coordinate conversion in OpenGLES an easy task?