NVIDIA 2D Image And Signal Performance Primitives (NPP)
Version 11.1.0.*

Primitives for computing the statistical properties of an image. More...
Modules  
Sum  
Primitives for computing the sum of all the pixel values in an image. 

Min  
Primitives for computing the minimal pixel value of an image. 

MinIndx  
Primitives for computing the minimal value and its indices (X and Y coordinates) of an image. 

Max  
Primitives for computing the maximal pixel value of an image. 

MaxIndx  
Primitives for computing the maximal value and its indices (X and Y coordinates) of an image. 

MinMax  
Primitives for computing both the minimal and the maximal values of an image. 

MinMaxIndx  
Primitives for computing the minimal and the maximal values with their indices (X and Y coordinates) of an image. 

Mean  
Primitives for computing the arithmetic mean of all the pixel values in an image. 

Mean_StdDev  
Primitives for computing both the arithmetic mean and the standard deviation of an image. 

Image Norms  
Primitives for computing the norms of an image, the norms of difference, and the relative errors of two images. 

DotProd  
Primitives for computing the dot product of two images. 

CountInRange.  
Primitives for computing the amount of pixels that fall into the specified intensity range. 

MaxEvery  
Primitives for computing the maximal value of the pixel pair from two images. 

MinEvery  
Primitives for computing the minimal value of the pixel pair from two images. 

Integral  
Primitives for computing the integral image of a given image. 

SqrIntegral  
Primitives for computing both the integral and the squared integral images of a given image. 

RectStdDev  
Primitives for computing the stansdard deviation of the integral images. 

HistogramEven  
Primitives for computing the histogram of an image with evenly distributed bins. 

HistogramRange  
Primitives for computing the histogram of an image within specified ranges. 

Image Proximity  
Primitives for computing the proximity measure between a source image and a template image. 

Image Quality Index  
Primitives for computing the image quality index of two images. 

MaximumError  
Primitives for computing the maximum error between two images. 

AverageError  
Primitives for computing the average error between two images. 

MaximumRelativeError  
Primitives for computing the maximum relative error between two images. 

AverageRelativeError  
Primitives for computing the average relative error between two images. 

IQA  
Primitives for computing the image quality between two images, such as MSE, PSNR, SSIM, and MSSSIM. 

Primitives for computing the statistical properties of an image.
Some statistical primitives also require scratch buffer during the computation. For details, please refer to Scratch Buffer and Host Pointer.
These functions can be found in the nppist library. Linking to only the sublibraries that you use can significantly save link time, application load time, and CUDA runtime startup time when using dynamic libraries.
Common parameters for nppiGetBufferHostSize functions include:
oSizeROI  RegionofInterest (ROI). 
hpBufferSize  Required buffer size. Important: hpBufferSize is a host pointer. Scratch Buffer and Host Pointer. 
nppStreamCtx  Application Managed Stream Context. 