Classification of herbs plant diseases via hierachical dynamic artificial neural network after image removal using kernel regression framework

Abdullah, Lili Nurliyana and Khalid, Fatimah and Borhan, N.M. (2011) Classification of herbs plant diseases via hierachical dynamic artificial neural network after image removal using kernel regression framework. International Journal on Computer Science and Engineering, 3 (1). pp. 15-20. ISSN 0975-3397

Full text not available from this repository.


When herbs plants has disease, they can display a range of symptoms such as colored spots, or streaks that can occur on the leaves, stems, and seeds of the plant. These visual symptoms continuously change their color, shape and size as the disease progresses. Once the image of a target is captured digitally, a myriad of image processing algorithms can be used to extract features from it. The usefulness of each of these features will depend on the particular patterns to be highlighted in the image. A key point in the implementation of optimal classifiers is the selection of features that characterize the image. Basically, in this study, image processing and pattern classification are going to be used to implement a machine vision system that could identify and classify the visual symptoms of herb plants diseases. The image processing is divided into four stages: Image Pre-Processing to remove image noises (Fixed-Valued Impulse Noise, Random-Valued Impulse Noise and Gaussian Noise), Image Segmentation to identify regions in the image that were likely to qualify as diseased region, Image Feature Extraction and Selection to extract and select important image features and Image Classification to classify the image into different herbs diseases classes. This paper is to propose an unsupervised diseases pattern recognition and classification algorithm that is based on a modified Hierarchical Dynamic Artificial Neural Network which provides an adjustable sensitivity-specificity herbs diseases detection and classification from the analysis of noise-free colored herbs images. It is also to proposed diseases treatment algorithm that is capable to provide a suitable treatment and control for each identified herbs diseases.

Item Type:Article
Keyword:HDNN; Bayesian algorithm; Fixed-valued impulse noise; Random-valued impulse noise and gaussian noise
Subject:Bayesian statistical decision theory
Subject:Neural networks (Computer science)
Faculty or Institute:Faculty of Computer Science and Information Technology
Publisher:Engg Journals Publications
ID Code:13603
Deposited By: Umikalthom Abdullah
Deposited On:29 Mar 2012 02:04
Last Modified:30 Oct 2012 08:30

Repository Staff Only: Edit item detail

Universiti Putra Malaysia Institutional Repository

Universiti Putra Malaysia Institutional Repository is an on-line digital archive that serves as a central collection and storage of scientific information and research at the Universiti Putra Malaysia.

Currently, the collections deposited in the IR consists of Master and PhD theses, Master and PhD Project Report, Journal Articles, Journal Bulletins, Conference Papers, UPM News, Newspaper Cuttings, Patents and Inaugural Lectures.

As the policy of the university does not permit users to view thesis in full text, access is only given to the first 24 pages only.