94
edits
Changes
→The parallelizing Express
We plan to parallelize the "ColorTransfer" program as it yields better theoretical speedup of the execution of the whole task in comparison to Sudoku and Sorted Array Processing.
== Assignment 2 ==
Description
Removing CPU Bottleneck
Removing the old CPU bottleneck in the byteCipher function:
<pre>
// Multiplication of matrix and vector
Color3d operator *(const cv::Mat& M, Color3d& v) {
Color3d u = Color3d();
for(int i=0; i<3; i++) {
u(i) = 0.0;
for(int j=0; j<3; j++) {
u(i) += M.at<double>(i, j) * v(j);
}
}
return u;
}
</pre>