NON CONTACT MEASUREMENT OF VERNONIA AMYGDALINA BITTERNESS FOR SCRUBBING MACHINE CONTROL

ABSTRACT

The study which we carried out whose topic is: “NON CONTACT MEASUREMENT OF VERNONIA AMYGDALINA BITTERNESS FOR SCRUBBING MACHINE CONTROL”, relates to the automatic washing of the leaves of V Amyg-dalina.

Therst part of this study shows that beyond nutritional qualities of the leaves of this plant, it is also used in the treatment of certain diseases. Its weak rate of consumption is with its bitterness a _rst study aiming at bringing a solution to this problem by the approach: “Global Image Characterization” was carried out last year by student SEN KEMBE Adel.

Unfortunately, this study was not completed. Our contribution with the resolution of this problem is the determination of the level of bitterness of bitter juice of the leaves of V Amygdalina in order to automate its system of washing.

The approach: “Region Image Characterization” that we deployed passes by the segmentation in two areas distinct (Juice and Wreck) of the images of the bitter juices of the leaves of V Amygdalina, then by the transformation of these areas into interpretable information by feature extraction.

In end, a correlation between extracted information and those obtained at the laboratory by our predecessor generate functions which the ow chart uses to determine the level of bitterness.

The analysis of the results of the two methods shows that the method by approach “Region Image Characterization” applied to the juice of the image is the best because error of measurement lower than the others and that it is more reliable with the magnication 10X.

Key words: Bitter leaves, images, image processing, segmentation, automated control unit.