#Rotating image in word software#
#Rotating image in word code#

An example in this chapter also utilizes OpenCL’s C++ wrapper.

Ĭhapter 4 dives into OpenCL programming examples, including histogram, image rotation, and convolution, and demonstrates some of the OpenCL 2.0 features, such as pipe memory objects.

The chapter includes a look into programming your first OpenCL application.

Ĭhapter 3 presents an introduction to OpenCL, including the host API and the OpenCL C language.

This chapter also covers the concepts of multicore and throughput-oriented systems, as well as advances in heterogeneous architectures. Different styles of architectures, including SIMD and very long instruction word, are discussed.

Ĭhapter 2 presents some of the architectures that support OpenCL, including CPUs, GPUs, and APUs.

#Rotating image in word software#

The chapter covers concurrency, threads, and different grains of parallelism: many of the fundamentals of parallel software development.

#Rotating image in word code#

For this example’s host code, we can reuse a substantial amount of code from the previous matrix multiplication example, including the code that will create the context and the command queue.Ĭhapter 1 (this chapter) introduces many concepts related to the development of parallel algorithms and software. The main steps for the host code are similar to those in Figure 4.2. In this case, because we rotating about the originĭest_data = src_data Īs seen in the previous kernel code, image rotation is an embarrassingly parallel problem, in which each resulting pixel value is computed independently. Read (ix,iy) src_data and store at (xpos,ypos) in Int ypos = (int)(yOff*cosTheta − xOff*sinTheta + y0) Int xpos = (int)(xOff*cosTheta + yOff*sinTheta + x0) Calculate location of data to move into (ix,iy) Work-item gets its index within index space _global float* dest_data, _global float* src_data,įloat sinTheta, float cosTheta) //Rotation Parameters For this reason, we need to check the bounds of the calculated output coordinates. When an image is rotated, the new locations of some pixels may be outside the image if the input and output image sizes are the same (see Figure 4.3, in which the corners of the input would not have fit within the resultant image). The image rotation example is a good example of an input decomposition, meaning that an element of the input (in this case, an input image) is decomposed into a work-item.