Soundous Belkacemi

Work place: Department of computer science, University Hadj Lakhdar, Batna, Algeria



Research Interests: Image Processing, Image Manipulation, Image Compression, Computer Vision, Computer systems and computational processes


Soundous Belkacemi, female, is Ph.D. student at the Department of computer science, University Hadj Lakhdar, Batna, (Algeria). Her research interests include computer vision, image processing.

Author Articles
Plants Leaves Images Segmentation Based on Pseudo Zernike Moments

By Ali Behloul Soundous Belkacemi

DOI:, Pub. Date: 8 Jun. 2015

Leaves images segmentation is an important task in the automated plant identification. Images leaf segmentation is the process of extracting the leaf from its background, which is a challenging task. In this paper, we propose an efficient and effective new approach for leaf image segmentation, we aim to separate the leaves from the background and from their shadow generated when the photo was taken. The proposed approach calculates the local descriptors for the image that will be classified for the separation of the different image's region. We use Pseudo Zernike Moments (PZM) as a local descriptor combined with K-means algorithm for clustering. The efficient of PZM for features extraction lead to very good results in very short time. The validation tests applied on a variety of images, showed the ability of the proposed approach for segmenting effectively the image. The results demonstrate a real improvement compared to those of new existing segmentation method.

[...] Read more.
Other Articles