Computational Physicshttp://ir.haramaya.edu.et//hru/handle/123456789/2152024-03-28T18:12:51Z2024-03-28T18:12:51ZAUTOMATIC MAIZE VARIETIES IDENTIFICATION USING ARTIFICIAL NEURAL NETWORKIMAGE CLASSIFICATION TECHNIQUESGetachew GobenaGetachew Abebe (PhD)http://ir.haramaya.edu.et//hru/handle/123456789/69952023-11-24T07:16:41Z2023-11-01T00:00:00ZAUTOMATIC MAIZE VARIETIES IDENTIFICATION USING ARTIFICIAL NEURAL NETWORKIMAGE CLASSIFICATION TECHNIQUES
Getachew Gobena; Getachew Abebe (PhD)
Maize is an economically important crop, which is contributing the highest of all food, industrial product and animal feeding revenues in Ethiopia. It is also the major cash crop of our country and farmers of the maize producing regions. Therefore, maize has a unique quality that fetches premium prices in the world market. However, to exploit this good opportunity the distinct unique shape, color and size of maize need to be maintained and improved. The first step to improve the crop is to understand the exiting variability in the genotype/accession. Therefore, in this research work, a digital image processing technique was used to classify maize seed based on their Color, morphological, and textural features. Artificial neural network (ANN) was used to classify maize varieties that are grown under different agronomic management and to check the maize seeds, which are grown under the same agronomic management. Four classification set-ups were used, which are classification based on color features, morphology features, texture features and combination of color, texture and morphology features. For the analysis, 12 colors, 6- morphological and 12 textural features, totally 30 features were extracted from images of maize grains. Four experimental classification setups were applied. Setups 1, 2, 3 and 4 are classification designs by using color, morphology, texture and combination of morphology and color features, respectively. For all setups 160 images were used as inputs of the ANN. From this datasets, 70%(112) images were used for training the ANN while the remaining 30%(48) images were used for both testing and validation. The accuracy of classification using color, feature 99.4%, morphological feature was 87.5%, texture features was 100% and combination of morphological and color features was 87.5% for maize seeds.
68
2023-11-01T00:00:00ZAUTOMATIC MAIZE VARIETIES IDENTIFICATION USING ARTIFICIAL NEURAL NETWORKIMAGE CLASSIFICATION TECHNIQUESGetachew Gobena WabulchoGetachew Abebe (PhDhttp://ir.haramaya.edu.et//hru/handle/123456789/69582023-11-24T05:58:21Z2023-11-01T00:00:00ZAUTOMATIC MAIZE VARIETIES IDENTIFICATION USING ARTIFICIAL NEURAL NETWORKIMAGE CLASSIFICATION TECHNIQUES
Getachew Gobena Wabulcho; Getachew Abebe (PhD
Maize is an economically important crop, which is contributing the highest of all food, industrial product and animal feeding revenues in Ethiopia. It is also the major cash crop of our country and farmers of the maize producing regions. Therefore, maize has a unique quality that fetches premium prices in the world market. However, to exploit this good opportunity the distinct unique shape, color and size of maize need to be maintained and improved. The first step to improve the crop is to understand the exiting variability in the genotype/accession. Therefore, in this research work, a digital image processing technique was used to classify maize seed based on their Color, morphological, and textural features. Artificial neural network (ANN) was used to classify maize varieties that are grown under different agronomic management and to check the maize seeds, which are grown under the same agronomic management. Four classification set-ups were used, which are classification based on color features, morphology features, texture features and combination of color, texture and morphology features. For the analysis, 12 colors, 6- morphological and 12 textural features, totally 30 features were extracted from images of maize grains. Four experimental classification setups were applied. Setups 1, 2, 3 and 4 are classification designs by using color, morphology, texture and combination of morphology and color features, respectively. For all setups 160 images were used as inputs of the ANN. From this datasets, 70%(112) images were used for training the ANN while the remaining 30%(48) images were used for both testing and validation. The accuracy of classification using color, feature 99.4%, morphological feature was 87.5%, texture features was 100% and combination of morphological and color features was 87.5% for maize seeds.
68
2023-11-01T00:00:00ZAUTOMATIC CLASSIFICATION OF SEVERITY STAGES OF PLAQUE PSORIASIS BY USING IMAGE PROCESSINGEFTU FEYUMAGetachew Abebe (PhD)http://ir.haramaya.edu.et//hru/handle/123456789/69182023-11-22T08:48:42Z2023-11-01T00:00:00ZAUTOMATIC CLASSIFICATION OF SEVERITY STAGES OF PLAQUE PSORIASIS BY USING IMAGE PROCESSING
EFTU FEYUMA; Getachew Abebe (PhD)
Psoriasis is a common skin disease affecting all ages and sexes and is characterized by itchy, scaly
patches most commonly on the knees, elbows, trunk, and scalp. The most common type is plaque
psoriasis, which causes raised, red patches on the skin that are covered with a silvery-white
buildup of dead skin cells, based on its severity which has three stages: mild, moderate, and severe.
Imaging technology has revolutionized diagnosis and treatment. Thus, this study is aimed at an
automatic classification of the severity stages of plaque psoriasis by using the Artificial Neural
Network technique. In total, 200 sample plaque psoriasis images were taken from Bisidimo
General Hospital, Oromia region, and Yem Dermatology and Venereology clinic, Harar. The
sample images were acquired using a Samsung Galaxy A14 mobile camera with (1080 × 2408)
resolution and loaded or imported to a computer. The imported images were pre-processed like
converting from RGB to gray-scale image and enhanced using CLAHE to increase the visibility of
the plaque psoriasis images. Then, thresholding was used to segment the plaque psoriasis-diseased
images, and the morphological feature was extracted from each image. Based on the
morphological feature, the arrangement for the input is carefully divided so that 70% for training
15% for validation, and 15% for testing sets. Artificial Neural Network (ANN) was used to classify
the plaque psoriasis skin disease severity stages as mild, moderate, and severe. The performance
measure of the classifier accuracy was computed. It was observed that the proposed scheme with
ANN classifier outperformed by giving 90% accuracy and a 10% probability of misclassification
error to classify plaque psoriasis as mild, moderate, or severe. All the techniques were
implemented through MATLAB R2018a platform and the design and implementation of these
image processing techniques help easy detection of plaque psoriasis skin diseases.
59
2023-11-01T00:00:00ZAUTOMATIC CLASSIFICATION OF TURCICUM LEAF BLIGHT AND COMMON RUST DISEASES ON MAIZE USING ARTIFICIAL NEURAL NETWORKLukas DanielGetachew Abebe (PhD)Prof. MashillaDejene (PhD)http://ir.haramaya.edu.et//hru/handle/123456789/67152023-11-02T06:28:37Z2023-06-01T00:00:00ZAUTOMATIC CLASSIFICATION OF TURCICUM LEAF BLIGHT AND COMMON RUST DISEASES ON MAIZE USING ARTIFICIAL NEURAL NETWORK
Lukas Daniel; Getachew Abebe (PhD); Prof. MashillaDejene (PhD)
Maize is an important crop, but it is vulnerable to plant diseases such as maize turcicum leaf
blight and common rust. Traditional methods to identify and classify maize leaf diseases have
drawbacks. Therefore, this study focused on developing a maize leaf disease detection and
classification algorithm using an artificial neural network technique. The maize leaf image
samples were taken from Haramaya University's glass house "Raree" Research Station using
random sampling techniques over three rounds. A total of 396 images were acquired using a
digital camera with high resolution and imported into a computer. The RGB images were
converted into CIELAB colour space, and the detection of turcicum leaf blight and common
rust was performed using k-means clustering to estimate the disease of the leaves. Then,
Otsu’s thresholding was used to segment the maize diseased leaf images, and morphological
features were extracted from each image. For the recognition and classification analysis, 4
morphological and 6 color features, totaling 10 features, were extracted from each image.To
evaluate the recognition and classification accuracy, from the total data set of 396 images,
70% were used for training, 15% were used for validation, and 15% were used for testing the
model. Based on morphological and color features, an artificial neural network (ANN) was
used to classify the maize leaf diseases.The performance measures of the classifier, such as
sensitivity, specificity, and accuracy, were computed. It was observed that the proposed
scheme with an ANN classifier outperformed the competition, giving 94.19% accuracy, 96.6%
sensitivity, and 92.2% specificity for morphological features, whereas the performance of the
classifier for color features were 93.5% accuracy, 92.6% sensitivity, and
94.2% specificity.Based on the experimental results using morphology and color features, the
morphological feature performed better than the color feature. All the techniques were
implemented through the MATLAB R2018a platform, and the design and implementation of
these image processing techniques help with the easy detection of plant diseases.
90
2023-06-01T00:00:00Z