
Intel® Integrated Performance Primitives for Image Processing
The second operational domain of Intel IPP is primitive functions for
processing "images", which is an umbrella term for twodimensional
data. Often Intel IPP for image processing are referred to as Intel IPPI
functions, by the prefix that the function name carries. The functions
of the image processing domain mostly are designed to perform operations
with images, video, and video coding, for example, raster operations (filters,
interpolation, geometric operations), MPEG2, MPEG4 and other video codecs,
JPEG, or computer vision.
List
of functions 
Description 
List of Functions 
Perform image data
manipulation, exchange and initialization operations.

Modify pixel values
of an image buffer using arithmetic or logical operations.
This group also includes functions that perform image
compositing based on opacity (alphablending).

Perform color
space conversion. The following color space conversions
are supported:
 Color models conversion
 Conversion from color to gray scale
 Reduction from high bit resolution color to low
bit resolution color
 Conversion from pixelorder to planar format and
vice versa
 Color twist
 Gamma correction.
All Intel IPP color conversion functions perform
point operations on pixels of the source image. For
a given destination pixel, the resultant channel values
are computed using channel values of the corresponding
source pixel only, and not involving any neighborhood
pixels. Thus, the rectangular region of interest (ROI)
used in function operations may extend to the size
of the whole image.

Perform related
operations on the pixelbypixel basis.

Perform morphological
erosion and dilation operations on images. Generally,
the erosion and dilation smooth the boundaries of objects
without significantly changing their area. Both operations
use either a symmetric 3x3 mask, a userdefined rectangular
mask, or a structuring element.

Perform linear
and nonlinear filtering operations on an image. Filtering
can be used in a variety of image processing operations;
for example, edge detection, blurring, noise removal,
and feature detection.

Perform linear
transform operations on an image buffer. These operations
include Fast Fourier Transform (FFT), Discrete Fourier
Transform (DFT), and Discrete Cosine Transform (DCT).


Click
one of the function types to the left or right to see a description
here. 
Compute the
following statistical parameters of an image:
 sum and mean of pixel values
 intensity histogram of pixel values
 minimum and maximum pixel values
 spatial and central moments of order 0 to 3
 the infinity, L1, and L2 norms of
the image pixel values and of the differences between
pixel values of two images
 relative error values for the infinity, L1,
and L2 norms of differences between pixel
values of two images.
 universal image quality index
 proximity measures of an image and a template
(another image).

Perform geometric
operations of resizing, rotating, warping and remapping
an image.

Perform the multiresolution analysis, which is a
better alternative to windowing and discrete Fourier
analysis techniques.
On the one hand, the forward twodimensional wavelet
transform may be considered as a decomposition of
an image on the base of functions bounded or localized
in space; and on the other, the wavelet transforms
are related to subband filtering and resampling.
Intel IPP for image processing contains onelevel
discrete wavelet decomposition and reconstruction
functions. It also provides the necessary interface
for initialization and deallocation of the transform
context structure.

Perform filtering,
feature detection, distance transform, flood fill, Gaussian
and Laplacian image pyramid generation and reconstruction,and
motion template functionality in computer vision.

Prepare data and perform still image compression
and coding in accordance with JPEG, lossless JPEG,
and JPEG2000 standards. These functions can be roughly
grouped as follows:
Functions for JPEG coding:
 Color conversion functions
 Combined color conversion functions
 Quantization functions
 Combined DCT functions
 Level shift functions
 Sampling functions
 Planartopixel and pixeltoplanar conversion
functions
 Huffman codec functions
Functions for JPEG2000 coding:
 Wavelet transform functions
 JPEG2000 entropy coding and decoding functions
 Component transform functions
Lossless JPEG functions

Perform encoding
and decoding of video data according to MPEG1, MPEG2,
MPEG4, DV, H.263, and H.264 standards. These functions
have a convenient interface and present an appropriate
solution for both encoding and decoding pipeline and,
as all other parts of Intel IPP, are intended for the
development of highperformance crossplatform code.




