KMITL

Permanent URI for this communityhttps://dspace.kmitl.ac.th/handle/123456789/1

Browse

Search Results

Now showing 1 - 10 of 22

Some of the metrics are blocked by your
consent settings
Black Gram Plant Nutrient Deficiency Classification in Combined Images Using Convolutional Neural Network
(2020-03-01)
Myo Han, Kadipa Aung
;
Watchareeruetai, Ukrit
Plant nutrient deficiency classification is vital for the agricultural industry to improve both the qualities and the quantities of crops. Computer vision and deep learning technologies, especially convolutional neural networks, perform an essential role in agricultural and biological sectors to solve the various kinds of complex problems. In this paper, we conducted the classification of the complete nutrient and six types of nutrient deficiency of black gram over the combined images of old leaf and young leaf. We found that the combined image supports more useful information than a single image. We accomplished the feature extraction process by taking the advantages of the deep pre-trained model to extract the features from the image automatically. Extracted features from ResNet50 deep pre-trained model are fed into three different classifiers as the input: (1) logistic regression, (2) support vector machine and (3) multilayer perceptron and compared the performance of these models. The multilayer perceptron models achieved superior performance than support vector machine and logistic regression by the accuracy of 88.33 %.
Some of the metrics are blocked by your
consent settings
Classification of nutrient deficiency in black gram using deep convolutional neural networks
(2019-07-01)
Han, Kadipa Aung Myo
;
Watchareeruetai, Ukrit
This paper investigates the use of various deep convolutional neural networks (CNNs) with transfer learning to identify nutrient deficiencies from a leaf image. Experiments were conducted with a dataset containing 4,088 images of black gram (Vigna mungo) leaves grown under seven different treatments, i.e., complete nutrient treatment and six nutrient deficiency treatment, including calcium (Ca), iron (Fe), magnesium (Mg), nitrogen (N), potassium (K), and phosphorus (P) deficiencies. Experimental results indicate that a deep CNN model known as ResNet50 was the best among all experimented models with a test accuracy of 65.44% and a F-measure of 66.15%. In addition, We found that the ResNet50 model obviously outperformed a block-based method and the human performance reported in a literature.
Some of the metrics are blocked by your
consent settings
An improvement of extreme learning machine using subclass clustering
(2018-07-02)
Watchareeruetai, Ukrit
;
Jiramaneepinit, Boonnithi
Extreme learning machine (ELM) is an extremely fast learning algorithm proposed for a single-hidden-layer feed-forward neural network (SLFN). ELM projects a set of training instances into a random feature space, and then analytically calculates the weight matrix connecting between the hidden layer and the output layer, leading to a very fast learning speed. This paper proposes an improved version of ELM, named clustering-ELM, that assigns a subclass to each training instances and learns for a weight matrix that projects random features into subclass. In the prediction step, the responses from output nodes of the same class are integrated into one using maximum function. Experimental results conducted on various benchmark datasets reveal a promising performance of the proposed clustering-ELM, compared to the standard ELM.
Some of the metrics are blocked by your
consent settings
Iterative extreme learning machine
(2018-07-02)
Jiramaneepinit, Boonnithi
;
Watchareeruetai, Ukrit
This paper proposes a simple but effective method to improve the generalization performance of extreme learning machine (ELM), which is an extremely fast learning method for a single-hidden-layer feedforward neural network (SLFN). The proposed method adopts an online sequential learning technique to update the output weight matrix of a learned SLFN by using misclassified training samples. As the process of updating these weights could be iteratively performed, the proposed method is named iterative ELM (I-ELM). The proposed I-ELM was evaluated on three datasets, including MNIST, Small NORB, and CIFAR-10, and compared with the standard ELM. Experimental results indicate that by using only a few iterations, the proposed I-ELM could effectively improve the generalization performance of SLFNs.
Some of the metrics are blocked by your
consent settings
Identification of Plant Nutrient Deficiencies Using Convolutional Neural Networks
(2018-07-02)
Watchareeruetai, Ukrit
;
Noinongyao, Pavit
;
Wattanapaiboonsuk, Chaiwat
;
Khantiviriya, Puriwat
;
Duangsrisai, Sutsawat
A novel image analysis method for identifying nutrient deficiencies in plant based on its leaf is proposed. First, the proposed method divides an input leaf image into small blocks. Second, each block of leaf pixels is fed to a set of convolutional neural networks (CNNs). Each CNN is specifically trained for a nutrient deficiency and is utilized to decide if a block is presenting any symptom of the corresponding nutrient deficiency. Next, the responses from all CNNs are integrated to produce a single response for the block using a winner-take-all strategy. Finally, the responses from all blocks are integrated into one using a multi-layer perceptron to produce a final response for the whole leaf. Validation of the proposed method was performed on a set of black gram (Vigna mungo) plants grown under nutrient-controlled environments. Five types of deficiencies, i.e., Ca, Fe, K, Mg, and N deficiencies, and a group of plants with complete nutrients were studied. A dataset consisting of 3,000 leaf images was collected and used for experimentation. Experimental results indicate the superiority of the proposed method over trained humans in nutrient deficiency identification.
Some of the metrics are blocked by your
consent settings
An extreme learning machine based pretraining method for multi-layer neural networks
(2018-07-02)
Noinongyao, Pavit
;
Watchareeruetai, Ukrit
One approach in training a deep neural network to perform effectively is to do unsupervised pretraining on each layer, followed by fine-tuning the whole network. A common way is to train an unsupervised model of neural network such as restricted Boltzmann machines or autoencoders and stack them on top of another. Although these unsupervised pretraining approaches yield good performance, relying on back-propagation, due to iterative learning process, they still suffer from a long pretraining time. Extreme learning machine (ELM) is an analytical training approach which is extremely fast and gives a solution with a good generalization performance. In this paper, we apply a new ELM based unsupervised learning, named backward ELM based autoencoder (BELM-AE), to pretrain each layer of a neural network before using a back-propagation based learning algorithm to fine-tune the whole network. Experimental results show that the new pretraining method requires significantly shorter training time and also yields better testing performance on various datasets.
Some of the metrics are blocked by your
consent settings
Detection of cotton wool for diabetic retinopathy analysis using neural network
(2017-12-13)
Bui, Toan
;
Maneerat, Noppadol
;
Watchareeruetai, Ukrit
This paper presents an automatic segmentation method used to detect cotton wool spots in the retinal images for diabetic retinopathy disease. An early detection of cotton wool is important to prevent the dangerous damage which may cause blindness and vision loss. A preprocessing is applied to enhance image quality followed by optic disc removal. A feature extraction method is used to take useful elements from the image for increasing accuracy in classification step. A neural network model is employed for learning task and tested by k-fold cross validation. Our approach is evaluated by ground truth on DIARETDB1 public data. The result shows that cotton wool can be segmented by this method with 85.9% in sensitivity, 84.4% in specificity, and 85.54% in accuracy.
Some of the metrics are blocked by your
consent settings
Image analysis algorithms for vehicle color recognition
(2017-10-19)
Tilakaratna, Damitha S.B.
;
Watchareeruetai, Ukrit
;
Siddhichai, Supakorn
;
Natcharapinchai, Nattachai
This work discusses about the implementation of a vehicle color recognition system to be used with the vehicle license plate recognition system. Because of the complexity of Thai alphabet, current license plate recognition system fails in correctly recognizing vehicles that have the same number but different Thai characters. Also the current system cannot identify illegal number plates that are being used. This color recognition system will help to resolve these problems and increase the accuracy of the vehicle recognition system. In addition, this system will provide a wide range of color classification which includes 13 colors including white, silver and gray colors. Color recognition is done by two methods: one with machine learning and one without machine learning. A best accuracy of 87.52% is given when using SVM to classify colors.
Some of the metrics are blocked by your
consent settings
Separation of abnormal regions on black gram leaves using image analysis
(2017-09-05)
Noinongyao, Pavit
;
Watchareeruetai, Ukrit
;
Khantiviriya, Puriwat
;
Wattanapaiboonsuk, Chaiwat
;
Duangsrisai, Sutsawat
This paper proposes an image analysis method for separating abnormal regions caused by nutrient deficiencies on plants' leaves. The proposed method analyzes a histogram of normal leaves' colors to identify abnormalities on leaves. It can be divided into three main steps. Firstly, color features of leaf region in an input image are computed. Secondly, for each pixel, its color features are compared to the corresponding bin in the histogram to determine whether the pixel is abnormal. Finally, a post-processing technique is then applied to reduce noises in the result. Experiments have been conducted using black gram (Vigna mungo) leaves with five different nutrient deficiencies. The experimental results show that the proposed method can separate abnormal regions with an accuracy of above 90%.
Some of the metrics are blocked by your
consent settings
Shape recognition by using Scale Invariant Feature Transform for contour
(2017-09-05)
Rojanamontien, Mathara
;
Watchareeruetai, Ukrit
This paper proposes a novel shape feature extractor named Contour-SIFT along with a matching method that computes the similarity between two set of proposed descriptors. It allows a shape to be recognized based on automatically located outstanding local features on its contour, which are extracted from 1-D signal representations of different smoothing scales. The algorithm describes each local feature as a list of frequencies from curvature histogram, which is created from curve segment around each local position. The descriptors will give high similarity compared with a model descriptors of a similar shape. The algorithm has properties of image scaling-, translation-, and rotation-invariants. An experiment were conducted with 200 images from Flavia dataset for verification. The result of using the proposed algorithm is compared with the result of using CSS.

KMITL

Browse

Filters

Settings

Sort By

Results per page

Search Results