<p>The steadily increasing demand of wood forces sawmills to increase the efficiency and effectiveness of their equipments. The weak parts and the twist in wooden logs have been documented as the most common and crucial defect in sawn lumber.</p><p>In this thesis we are going to implement a program which is able to determine the weak parts in wooden logs. The implementation will be in two languages C++ and Matlab. Parts of the program are going to be implemented sometimes by C++ and sometimes by Matlab therefore different designs are going to be tested. The aim of this thesis is to check whether these designs will meet the real time bound of 10 m/s.</p><p>The result shows that there is a huge difference in time performance for the different designs. Therefore, different discretization levels were used in order to meet the deadline of 10m/s. We found that in order to get better speed one should calculate the matrix for the function <em>F </em>and the Jacobian function <em>J </em>by using C++ not by Matlab. Also we found that when we called functions from one language to another an extra time is added.</p>

The steadily increasing demand of wood forces sawmills to increase the efficiency and effectiveness of their equipments. The weak parts and the twist in wooden logs have been documented as the most common and crucial defect in sawn lumber. In this thesis we are going to implement a program which is able to determine the weak parts in wooden logs. The implementation will be in two languages C++ and Matlab. Parts of the program are going to be implemented sometimes by C++ and sometimes by Matlab therefore different designs are going to be tested. The aim of this thesis is to check whether these designs will meet the real time bound of 10 m/s. The result shows that there is a huge difference in time performance for the different designs. Therefore, different discretization levels were used in order to meet the deadline of 10m/s. We found that in order to get better speed one should calculate the matrix for the function F and the Jacobian function J by using C++ not by Matlab. Also we found that when we called functions from one language to another an extra time is added.