CUVI Features

CUVI SDK is an image processing kit written and developed on C-CUDA and C++ for over a period of ten years. The SDK includes 100+ image processing and computer vision functions that are plug-n-play for imaging applications in a plethora of verticals. Each function is hand crafted and highly optimized to run on CUDA supported GPUs. Many functions use our own proprietary approach to algorithms and processing to make the best out of hardware and input data.

Modules

CUVI comes in following image processing modules.

1. Arithmetic & Logical
2. Color Operations
3. Computer Vision
4. Data Exchange
5. Geometry Transforms
6. Image Filtering
7. Image Statistics
8. Image Transforms
9. Image Codecs

Function List

Arithmetic & Logical abs - Computes absolute value of each image pixel absDiff - Computes absolute difference between 2 images. add - Adds two images AND - Bitwise AND operation between two images. cbrt - Computes cube root of each image pixel ceil - Applies ceil function to each image pixel divide - Divides pixel values of two images dot exp - Computes exponential of each image pixel floor - Applies floor function to each image pixel inRange - Checks if image pixels lie between a range of values invert - Computes matrix inverse using LU factorization matMul - Performs matrix multiplication matMulTrans - Performs matrix multiplication with its transpose multiply - Multiplies pixel values of two images NOT - Inverts the bits of each pixel. OR - Bitwise OR operation between two images. pow - Raise each pixel to a specified power. round - Rounds each pixel to nearest integer value. SHL - Left shift each pixel of an image. SHR - Right shift each pixel of an image sqrt - Takes square root of each pixel of an image. subtract - Subtracts pixel values of two images. XOR - Logical bitwise XOR operation between two images. Color Operations adjust - Equivalent to MATLAB's imadjust autoColor - Fixes false colors at pixel level using a propriety approach borderMask - Masks image borders with custom intensity blackGammaLUT - Remaps image values according to 3 look up tables channelMix - Applies a color twist matrix to an image channelMerge - Merges R, G and B Channel to form a color image channelSplit - Splits a three channel image into R, G and B channels colorPick - Performs color segmentation of image by selecting a single color dehaze - Makes sense of images taken in a blizzard demosaic - Restores an RGB image from a CFA Bayer image demosaicHQLI - Restores an RGB using advanced hq debayering demosaicDFPD - Restores an RGB image from a CFA Bayer image using DFPD algorithm fpn - fixed point noise remove via camera black image FPNCorrection - fixed point noise removal via black and white images and gain gammaCorrect - Performs gamma correction of an image gray2rgb - Copy a grayscale image to each channel of an RGB image. histEq - Performs histogram equalization hsv2rgb - Converts image from HSV color space to RGB. imageBinary - Converts input image into a binary image based on threshold Lab2rgb - Converts image from CIE Lab color space to RGB. logTransform - Applies log transformation for contrast enhancement of an image lowlight - enhances the low intensity colors in a dark photo LUT - Remaps image values according to look up table rgb2gray - Converts an RGB image to gray scale rgb2hsv - Converts an RGB image to HSV color space rgb2Lab - Converts an RGB image to CIE Lab color space rgb2ycbcr - Converts an RGB image to YCbCr color space rgb2yuv - Converts RGB image to YUV color space ycbcr2rgb - Converts YCbCr image to an RGB image yuv2rgb - Converts YUV image to an RGB image. YUV422TOV210 - Converts YUV422 image to an V210 image. YUV444TOV210 - Converts YUV444 image to an V210 image. V210TOYUV444 - Converts V210 image to an YUV444 image. Computer Vision eigenValsVecs - Calculates eigen values and eigen vectors at each pixel in an image findTransformECC - Calculates geometric transform between two images in terms of the ECC criterion goodFeaturesToTrack - selects N best points on a image using KLT or Harris minEigenVal - Calculates the minimum eigen value at each pixel in an image opticalFlowHS - Calculates flow of each image pixel in two frames using Horn–Schunck method opticalFlowPyrLKDense - Calculates flow of each image pixel in two frames using pyramidal Lucas–Kanade method trackFeatures - Tracks input features from frame 1 onto frame 2 using KLT method focusStack - Stacks multiples photos with various focus into a single all focused image Data Exchange bitConversion - Converts image pixel values from one data type to another. convertScale - Converts image data type with optional scaling. generateRandomImage - Creates a new random image. getChannel - Extracts a channel of an image. padImage - Pads any side of the image with the value specified. rectangle - Draws a rectangle on the specified region of an image. threshold - Clamps image pixel values between specified thresholds. Geometry Transforms crop - Crops the desired portion of an image flipImage - Mirrors an image about a horizontal or vertical axis, or both resize - Resizes input image using specified interpolation rotate - Rotates an image about specified anchor RotateNoCrop - Rotates an image about specified anchor taper - Smartly extends image border transpose - Takes Transpose of the image warpAffine - Applies an affine transformation to an image. warpAffineBack - Applies an inverse affine transformation to an image. warpPerspective - Applies an affine transformation to an image. warpPerspectiveBack - Applies an inverse affine transformation to an image. Image Filtering imageFilter - Performs linear 1D and 2D image filtering imageGradients - Calculates X and Y gradients of image minFilter - Filters an image using a min filter maxFilter - Filters an image using a max filter medianFilter - Applies 2D median filter on an image underwaterFilter - Brings out original colors of underwater photography. Image Statistics calcHist - Computes the histogram of an 8 bit image. countInRange - Counts the number of pixels within the given intensity range max - Computes the maximum of image pixel values mean - Computes the mean of image pixel values meanStdDev - Computes the mean and standard deviation of image pixel values min - Computes the minimum of image pixel values minMax - Computes the minimum and maximum of image pixel values sum - Computes the sum of image pixel values Image Transforms fft2 - Computes the 2D Fast Fourier Transform of an image. fftGetProperty - Gets magnitude, phase, real and imaginary part of an FFT image with optional scaling. fftShift - Moves low frequency components to the center of an FFT image. haarFwd - Performs one-level wavelet decomposition of an image using haar basis haarInv - Performs one-level haar wavelet reconstruction of an image Image Codecs encode - Encodes image pixel array into JPEG. decode - Decodes a JPEG into an image pixel array.

CUVI Core

CUVI Core includes the following helper functionality:

CuviImage A C++ class to hold image data on the device. It has the following methods CuviImage(); - default constructor CuviImage(CuviSize size, Cuvi32s depth, Cuvi32s channels) - Main Constructor. throws exception in case of failure CuviImage(const CuviImage&) - copy constructor upload(const void* pSrcHost, const Cuvi32s srcPitch) - Copy image data from host to device. upload(const void* pSrcHost, const Cuvi32s srcPitch, const CuviStream& stream) - Asynchronously copy image data from host to device. upload(const void* pSrcHost, CuviRect srcROI, const Cuvi32s srcPitch, CuviPoint2D<int> dstOrigin) - Copy a chunk of image data from host to device. upload(const void* pSrcHost, CuviRect srcROI, const Cuvi32s srcPitch, CuviPoint2D<int> dstOrigin, const CuviStream& stream) - Asynchronously copy a chunk of image data from host to device. download(void* pDstHost, const Cuvi32s dstPitch) - Copy image data from device to host download(void* pDstHost, const Cuvi32s dstPitch, const CuviStream& stream) - Asynchronously copy image data from device to host download(void* pDstHost, CuviPoint2D<int> dstOrigin, const Cuvi32s dstPitch, CuviRect srcROI) - Copy a chunk of image data from device to host download(void* pDstHost, CuviPoint2D<int> dstOrigin, const Cuvi32s dstPitch, CuviRect srcROI, const CuviStream& stream) - Asynchronously copy a chunk of image data from device to host copyTo(CuviImage& dst, CuviPoint2D<int> dstOrigin, CuviRect srcROI, const CuviStream& stream = CuviStream()) - Copy image data (or an ROI of it) from one Cuvi Image to another. show(const std::string& title = "CUVI Image", Cuvi32s milliseconds = 0) - Render device image directly on the screen ~CuviImage() - Destructor. Frees Device Memory. throws exception in case of failure Timer CUVI provides a C++ Timer class that accurately profiles your application timing Start() - Starts the timer. Stop() - Stops the timer Reset() - Resets the timer GetElapsedTime() - Returns the elapsed time between Start and Stop calls in seconds DeviceProperties CuviDeviceProperties holds all the important properties of underlying CUDA device. Device management functions are present inside the namespace cuvi::device. Cuvi32s getDeviceCount() - Gives the count of devices installed in the machine Cuvi32s getCurentDevice() - Returns the ID of selected graphics card CuviStatus setCurrentDevice(const Cuvi32s ID) - Makes any device an active device CuviStatus getDeviceProperties(const Cuvi32s deviceID, CuviDeviceProperties& props) - Returns all the important properties of desired GPU