Complexity Analysis and Performance of the Madenda Filter Algorithm

There are many algorithms used to do edge detection. Algorithms on edge detection, in general, can be applied to detect the edge of the image that is affected by noise and not exposed to noise. However, it has not been able to filter for images with noise problems that also experience blurry effects. The Madenda filter algorithm develops from the Canny filter algorithm and Shen-Castan algorithm so that it can overcome this problem. This study aims to measure the time complexity of the Madenda filter algorithm and perform a performance evaluation on the filter. The analysis of the complexity of the Madenda filter algorithm uses the Big O notation. Whereas the performance evaluation is carried out with the help of MATLAB software (trial version). Based on the running time of the Madenda filter algorithm is obtained as follows from the results of time complexity calculations, it is obtained if the Madenda filter algorithm has the complexity of the O (n
²
). Based on the implementation of program-code implemented in the MATLAB programming language shows that the Madenda filter algorithm can do edge detection where images other than experiencing noise disturbances also experience blurry effects. But the condition that must be fulfilled is that the α value must be smaller and the blur controller parameter (β) approaches 1.