The Fast Multilevel Fuzzy Edge Detection of Blurry Image

Summary:

This paper proposes a method of edge lines in the image blur using a fast multilevel fuzzy edge detection. This method has two stages, the first image contrast improved by an average of fast multilevel fuzzy enhancement algorithm (FMFE). FMFE used in the threshold Q dividing the image into two regions, namely FO and FB. FO is an area with a pixel grayish High and FB area with pixel gray levels. Then calculated the average area FO and FB named AO and AB. Then transformed into a fuzzy area. Of the AO

and AB are used to obtain the degree of membership of each pixel. The second stage of extracting edge lines using a two-stage edge detection operator which identify candidates by margins of local characteristics and determine the pixel border that actually uses edge detection operator based extremum of the gradient.

advantages:

Can detect the edges of image blur.

weaknesses:

The first stage is used to improve image contrast with FMFE algorithm. The FMFE requires optimal threshold value Q. The Q divides the image into two parts, namely the pixel with the grayish pixels high and the low gray. Disadvantage if the Q not optimal, the results obtained are not optimal.

Suggestions:

Thresholding method can obtain the optimum value of Q eg thresholding using hierarchical cluster analysis.

Contributions:

Modify the thresholding method used to obtain the Q method thresholding using hierarchical cluster analysis.