/*
-----------------------------------------------------------------------------
This source file is part of OGRE
(Object-oriented Graphics Rendering Engine)
For the latest info, see http://www.ogre3d.org/
Copyright (c) 2000-2014 Torus Knot Software Ltd
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in
all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
THE SOFTWARE.
-----------------------------------------------------------------------------
*/
#ifndef _Image_H__
#define _Image_H__
#include "OgrePrerequisites.h"
#include "OgreCommon.h"
#include "OgrePixelFormat.h"
namespace Ogre {
/** \addtogroup Core
* @{
*/
/** \addtogroup Image
* @{
*/
enum ImageFlags
{
IF_COMPRESSED = 0x00000001,
IF_CUBEMAP = 0x00000002,
IF_3D_TEXTURE = 0x00000004
};
/** Class representing an image file.
The Image class usually holds uncompressed image data and is the
only object that can be loaded in a texture. Image objects handle
image data decoding themselves by the means of locating the correct
Codec object for each data type.
@par
Typically, you would want to use an Image object to load a texture
when extra processing needs to be done on an image before it is
loaded or when you want to blit to an existing texture.
*/
class _OgreExport Image : public ImageAlloc
{
friend class ImageCodec;
public:
/** Standard constructor.
*
* allocates a buffer of given size if buffer pointer is NULL.
*/
Image(PixelFormat format = PF_UNKNOWN, uint32 width = 0, uint32 height = 0, uint32 depth = 1,
uchar* buffer = NULL, bool autoDelete = true);
/** Copy-constructor - copies all the data from the target image.
*/
Image( const Image &img );
/**
* allocates a buffer of given size if needed
*
* - If the current allocation is equal to the requested size, this does nothing
* - Otherwise any current allocation is freed, and memory of specified size is allocated
*
* @see loadDynamicImage
*/
void create(PixelFormat format, uint32 width, uint32 height, uint32 depth = 1, uint32 numFaces = 1,
uint32 numMipMaps = 0);
/** Standard destructor.
*/
~Image();
/** Assignment operator - copies all the data from the target image.
*/
Image & operator = ( const Image & img );
/**
* sets all pixels to the specified colour
*
* format conversion is performed as needed
*/
void setTo(const ColourValue& col);
/** Flips (mirrors) the image around the Y-axis.
An example of an original and flipped image:
originalimg
00000000000
00000000000
00000000000
00000000000
00000000000
------------> flip axis
00000000000
00000000000
00000000000
00000000000
00000000000
originalimg
*/
Image & flipAroundY();
/** Flips (mirrors) the image around the X-axis.
An example of an original and flipped image:
flip axis
|
originalimg|gmilanigiro
00000000000|00000000000
00000000000|00000000000
00000000000|00000000000
00000000000|00000000000
00000000000|00000000000
*/
Image & flipAroundX();
/** Stores a pointer to raw data in memory. The pixel format has to be specified.
This method loads an image into memory held in the object. The
pixel format will be either greyscale or RGB with an optional
Alpha component.
The type can be determined by calling getFormat().
@note
Whilst typically your image is likely to be a simple 2D image,
you can define complex images including cube maps, volume maps,
and images including custom mip levels. The layout of the
internal memory should be:
- face 0, mip 0 (top), width x height (x depth)
- face 0, mip 1, width/2 x height/2 (x depth/2)
- face 0, mip 2, width/4 x height/4 (x depth/4)
- .. remaining mips for face 0 ..
- face 1, mip 0 (top), width x height (x depth)
.. and so on.
Of course, you will never have multiple faces (cube map) and
depth too.
@param data
The data pointer
@param width
Width of image
@param height
Height of image
@param depth
Image Depth (in 3d images, numbers of layers, otherwise 1)
@param format
Pixel Format
@param autoDelete
If memory associated with this buffer is to be destroyed
with the Image object. Note: it's important that if you set
this option to true, that you allocated the memory using OGRE_ALLOC_T
with a category of MEMCATEGORY_GENERAL to ensure the freeing of memory
matches up.
@param numFaces
The number of faces the image data has inside (6 for cubemaps, 1 otherwise)
@param numMipMaps
The number of mipmaps the image data has inside
@note
The memory associated with this buffer is NOT destroyed with the
Image object, unless autoDelete is set to true.
The size of the buffer must be numFaces * PixelUtil::getMemorySize(width, height, depth, format)
*/
Image& loadDynamicImage(uchar* data, uint32 width, uint32 height, uint32 depth, PixelFormat format,
bool autoDelete = false, uint32 numFaces = 1, uint32 numMipMaps = 0);
/// @overload
Image& loadDynamicImage(uchar* data, uint32 width, uint32 height, PixelFormat format)
{
return loadDynamicImage(data, width, height, 1, format);
}
/** Loads raw data from a stream. See the function
loadDynamicImage for a description of the parameters.
The size of the buffer must be numFaces * PixelUtil::getMemorySize(width, height, depth, format)
@note
Whilst typically your image is likely to be a simple 2D image,
you can define complex images including cube maps
and images including custom mip levels. The layout of the
internal memory should be:
- face 0, mip 0 (top), width x height (x depth)
- face 0, mip 1, width/2 x height/2 (x depth/2)
- face 0, mip 2, width/4 x height/4 (x depth/4)
- .. remaining mips for face 0 ..
- face 1, mip 0 (top), width x height (x depth)
.. and so on.
Of course, you will never have multiple faces (cube map) and
depth too.
*/
Image& loadRawData(const DataStreamPtr& stream, uint32 width, uint32 height, uint32 depth,
PixelFormat format, uint32 numFaces = 1, uint32 numMipMaps = 0);
/// @overload
Image& loadRawData(const DataStreamPtr& stream, uint32 width, uint32 height, PixelFormat format)
{
return loadRawData(stream, width, height, 1, format);
}
/** Loads an image file.
This method loads an image into memory. Any format for which
an associated ImageCodec is registered can be loaded.
This can include complex formats like DDS with embedded custom
mipmaps, cube faces and volume textures.
The type can be determined by calling getFormat().
@param
filename Name of an image file to load.
@param
groupName Name of the resource group to search for the image
@note
The memory associated with this buffer is destroyed with the
Image object.
*/
Image & load( const String& filename, const String& groupName );
/** Loads an image file from a stream.
This method works in the same way as the filename-based load
method except it loads the image from a DataStream object.
This DataStream is expected to contain the
encoded data as it would be held in a file.
Any format for which an associated ImageCodec is registered
can be loaded.
This can include complex formats like DDS with embedded custom
mipmaps, cube faces and volume textures.
The type can be determined by calling getFormat().
@param
stream The source data.
@param
type The type of the image. Used to decide what decompression
codec to use. Can be left blank if the stream data includes
a header to identify the data.
@see
Image::load( const String& filename )
*/
Image & load(const DataStreamPtr& stream, String type = BLANKSTRING );
/** Utility method to combine 2 separate images into this one, with the first
image source supplying the RGB channels, and the second image supplying the
alpha channel (as luminance or separate alpha).
@param rgbFilename Filename of image supplying the RGB channels (any alpha is ignored)
@param alphaFilename Filename of image supplying the alpha channel. If a luminance image the
single channel is used directly, if an RGB image then the values are
converted to greyscale.
@param groupName The resource group from which to load the images
@param format The destination format
*/
Image & loadTwoImagesAsRGBA(const String& rgbFilename, const String& alphaFilename,
const String& groupName, PixelFormat format = PF_BYTE_RGBA);
/** Utility method to combine 2 separate images into this one, with the first
image source supplying the RGB channels, and the second image supplying the
alpha channel (as luminance or separate alpha).
@param rgbStream Stream of image supplying the RGB channels (any alpha is ignored)
@param alphaStream Stream of image supplying the alpha channel. If a luminance image the
single channel is used directly, if an RGB image then the values are
converted to greyscale.
@param format The destination format
@param rgbType The type of the RGB image. Used to decide what decompression
codec to use. Can be left blank if the stream data includes
a header to identify the data.
@param alphaType The type of the alpha image. Used to decide what decompression
codec to use. Can be left blank if the stream data includes
a header to identify the data.
*/
Image& loadTwoImagesAsRGBA(const DataStreamPtr& rgbStream, const DataStreamPtr& alphaStream,
PixelFormat format = PF_BYTE_RGBA,
const String& rgbType = BLANKSTRING,
const String& alphaType = BLANKSTRING);
/** Utility method to combine 2 separate images into this one, with the first
image source supplying the RGB channels, and the second image supplying the
alpha channel (as luminance or separate alpha).
@param rgb Image supplying the RGB channels (any alpha is ignored)
@param alpha Image supplying the alpha channel. If a luminance image the
single channel is used directly, if an RGB image then the values are
converted to greyscale.
@param format The destination format
*/
Image & combineTwoImagesAsRGBA(const Image& rgb, const Image& alpha, PixelFormat format = PF_BYTE_RGBA);
/** Save the image as a file.
Saving and loading are implemented by back end (sometimes third
party) codecs. Implemented saving functionality is more limited
than loading in some cases. Particularly DDS file format support
is currently limited to true colour or single channel float32,
square, power of two textures with no mipmaps. Volumetric support
is currently limited to DDS files.
*/
void save(const String& filename);
/** Encode the image and return a stream to the data.
@param formatextension An extension to identify the image format
to encode into, e.g. "jpg" or "png"
*/
DataStreamPtr encode(const String& formatextension);
/** Returns a pointer to the internal image buffer at the specified pixel location.
Be careful with this method. You will almost certainly
prefer to use getPixelBox, especially with complex images
which include many faces or custom mipmaps.
*/
uchar* getData(uint32 x = 0, uint32 y = 0, uint32 z = 0)
{
assert((!mBuffer && (x + y + z) == 0) || (x < mWidth && y < mHeight && z < mDepth));
return mBuffer + mPixelSize * (z * mWidth * mHeight + mWidth * y + x);
}
/// @overload
const uchar* getData(uint32 x = 0, uint32 y = 0, uint32 z = 0) const
{
assert(mBuffer);
assert(x < mWidth && y < mHeight && z < mDepth);
return mBuffer + mPixelSize * (z * mWidth * mHeight + mWidth * y + x);
}
/// @overload
template T* getData(uint32 x = 0, uint32 y = 0, uint32 z = 0)
{
return reinterpret_cast(getData(x, y, z));
}
/// @overload
template const T* getData(uint32 x = 0, uint32 y = 0, uint32 z = 0) const
{
return reinterpret_cast(getData(x, y, z));
}
/** Returns the size of the data buffer in bytes
*/
size_t getSize() const { return mBufSize; }
/** Returns the number of mipmaps contained in the image.
*/
uint32 getNumMipmaps() const { return mNumMipmaps; }
/** Returns true if the image has the appropriate flag set.
*/
bool hasFlag(const ImageFlags imgFlag) const { return (mFlags & imgFlag) != 0; }
/** Gets the width of the image in pixels.
*/
uint32 getWidth(void) const { return mWidth; }
/** Gets the height of the image in pixels.
*/
uint32 getHeight(void) const { return mHeight; }
/** Gets the depth of the image.
*/
uint32 getDepth(void) const { return mDepth; }
/** Get the number of faces of the image. This is usually 6 for a cubemap, and
1 for a normal image.
*/
uint32 getNumFaces(void) const { return hasFlag(IF_CUBEMAP) ? 6 : 1; }
/** Gets the physical width in bytes of each row of pixels.
*/
size_t getRowSpan(void) const { return mWidth * mPixelSize; }
/** Returns the image format.
*/
PixelFormat getFormat() const { return mFormat; }
/** Returns the number of bits per pixel.
*/
uchar getBPP() const { return mPixelSize * 8;}
/** Returns true if the image has an alpha component.
*/
bool getHasAlpha() const;
/** Does gamma adjustment.
@note
Basic algo taken from Titan Engine, copyright (c) 2000 Ignacio
Castano Iguado
*/
static void applyGamma( uchar *buffer, Real gamma, size_t size, uchar bpp );
/**
* Get colour value from a certain location in the image. The z coordinate
* is only valid for cubemaps and volume textures. This uses the first (largest)
* mipmap.
*/
ColourValue getColourAt(uint32 x, uint32 y, uint32 z) const;
/**
* Set colour value at a certain location in the image. The z coordinate
* is only valid for cubemaps and volume textures. This uses the first (largest)
* mipmap.
*/
void setColourAt(ColourValue const &cv, uint32 x, uint32 y, uint32 z);
/**
* Get a PixelBox encapsulating the image data of a mipmap
*/
PixelBox getPixelBox(uint32 face = 0, uint32 mipmap = 0) const;
/// Delete all the memory held by this image, if owned by this image (not dynamic)
void freeMemory();
enum Filter
{
FILTER_NEAREST,
FILTER_LINEAR,
FILTER_BILINEAR = FILTER_LINEAR
};
/** Scale a 1D, 2D or 3D image volume.
@param src PixelBox containing the source pointer, dimensions and format
@param dst PixelBox containing the destination pointer, dimensions and format
@param filter Which filter to use
This function can do pixel format conversion in the process.
@note dst and src can point to the same PixelBox object without any problem
*/
static void scale(const PixelBox &src, const PixelBox &dst, Filter filter = FILTER_BILINEAR);
/** Resize a 2D image, applying the appropriate filter. */
void resize(ushort width, ushort height, Filter filter = FILTER_BILINEAR);
/// Static function to calculate size in bytes from the number of mipmaps, faces and the dimensions
static size_t calculateSize(uint32 mipmaps, uint32 faces, uint32 width, uint32 height, uint32 depth, PixelFormat format);
/// Static function to get an image type string from a stream via magic numbers
OGRE_DEPRECATED static String getFileExtFromMagic(DataStreamPtr stream);
private:
/// The width of the image in pixels
uint32 mWidth;
/// The height of the image in pixels
uint32 mHeight;
/// The depth of the image
uint32 mDepth;
/// The number of mipmaps the image contains
uint32 mNumMipmaps;
/// The size of the image buffer
size_t mBufSize;
/// Image specific flags.
int mFlags;
/// The pixel format of the image
PixelFormat mFormat;
uchar* mBuffer;
/// The number of bytes per pixel
uchar mPixelSize;
/// A bool to determine if we delete the buffer or the calling app does
bool mAutoDelete;
};
typedef std::vector ImagePtrList;
typedef std::vector ConstImagePtrList;
/** @} */
/** @} */
} // namespace
#endif