Garuda - Garba Rujukan Digital

On Reducing ShuffleNets’ Block for Mobile-based Breast Cancer Detection Using Thermogram: Performance Evaluation Rizka Ramadhana; Khairun Saddami; Khairul Munadi; Fitri Arnia
Indonesian Journal of Electrical Engineering and Informatics (IJEEI) Vol 10, No 4: December 2022
Publisher : IAES Indonesian Section

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52549/ijeei.v10i4.4062

In this paper, we proposed a reduced-block-Shufflenet (RB-ShuffleNet) for thermal breast cancer detection. RB-ShuffleNet is a modification of Shufflenet obtained by reducing blocks from the original architecture. The images for training and testing were obtained from Database for Mastology Research (DMR). First, we detected and cropped the image based on the region of interest (ROI), in which the ROI is determined by using the red intensity profile. Then, the ROI images were trained using RB-ShuffleNets. In the experiments, we built eight architectures, based on ShuffleNet, each with a different number of reduced blocks. The result showed that RB-Shufflenet with four reduced blocks had fewer than 50% of the learning parameters of the original Shufflenet, without compromising its performance. The RB-ShuffleNet with up to four reduced blocks could achieve 100% testing accuracy. Furthermore, The RB-ShuffleNets performed better than MobileNetV2 and resulted in higher accuracy when fed with ROI images. Due to its light structure and good performance, we recommend RB-ShuffleNet as mobile-based CNN model which is preferable to implement in breast cancer detection.

Fine Tuning CNN Pre-trained Model Based on Thermal Imaging for Obesity Early Detection Hendrik Leo; Fitri Arnia; Khairul Munadi
Jurnal Rekayasa Elektrika Vol 18, No 1 (2022)
Publisher : Universitas Syiah Kuala

Obesity is a complex disease that causes serious impact health, such as diabetes mellitus, cardiovascular disease, cancer, and stroke. An early obesity diagnosis/ detection method is required to prevent the increasing number of obese people. This study aims to: (i) fine-tune the pre-trained Convolutional Neural Network (CNN) models to build an early detection of obesity and (ii) evaluate the model performance in terms of classifying performance, computation speed, and learning performance. The thermal images acquisition procedure was conducted with 18 normal subjects and 15 obese subjects to build a thermal images dataset of obesity. Pre-trained CNN models: VGG19, MobileNet, ResNet152V, and DenseNet201 were modified and trained using the acquired dataset as the input. The training results show that the DenseNet201 model outperformed other models regarding classifying accuracy: 83.33 % and learning performances. At the same time, the MobileNet model outperformed other models in terms of computation speed with training elapsed time: 12 seconds/epoch. The proposed DenseNet201 model was suitable for implementation as an early screening system of obesity for health workers or physicians. Meanwhile, the proposed MobileNet model was suitable for mobile applications' early detection/diagnosis of obesity.

Identifikasi Tingkat Kematangan Kelapa Sawit Berbasis Pencitraan Termal Khusnul Azima; Khairul Munadi; Fitri Arnia; Maulisa Oktiana
Jurnal Rekayasa Elektrika Vol 15, No 1 (2019)
Publisher : Universitas Syiah Kuala

Indonesia is the biggest producer of palm oil (Elaeis guineenis jacq). The palm tree is a primary commodity that posses a high economic value. Palm oil must be considered in terms of quality to produce optimal and high-quality oil. Previously, the stipulation of the palm tree characterization used manual and visual image utilization method; it may have weaknesses due to the dependency of individual sorting and coruscation factor. Therefore, this research is aimed to improve the performance of the previous method in identifying the ripeness of palm tree based on thermal imaging. The excess of thermal imaging was not related to the coruscation since the level of ripeness was both determined by the temperature and colour. The detection method of this research deployed the colour-based features that are Dominant Colour Descriptor and Color Moment. The DCD and Color Moment was the input to the K-Nearest Neighbor (KNN) method. The percentage of identification rate was 89%, and the identification of oil palm maturity level using thermal imaging is more efficient because it is done without human intervention and does not depend on lighting assistance compared to manual method and method of using RGB visual images.

Pengukuran Badan Ikan Berupa Estimasi Panjang, Lebar, dan Tinggi Berdasarkan Visual Capture Raihan Islamadina; Nuriza Pramita; Fitri Arnia; Khairul Munadi; TWK Muhammad Iqbal
Jurnal Nasional Teknik Elektro dan Teknologi Informasi Vol 7 No 1: Februari 2018
Publisher : Departemen Teknik Elektro dan Teknologi Informasi, Fakultas Teknik, Universitas Gadjah Mada

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (1308.794 KB)

Currently, fish measurement process is done manually using the gauge which can cause inappropriate, ineffective result, and requires long time to finish, especially for a great amount of fish. Therefore, an automatic fish body measurement technique in the form of estimation of length, width, and height of fish based on visual capture is needed to facilitate fish body measurement to become more effective and efficient. This research uses five samples of fish in measuring length, width, and height manually to obtain the average data/value as the calibration of reference value for calculation process in the system and to be stored in the database. The stages begin with capturing fish image using a digital camera. Then, preprocessing stage was carried out to get the grayscale image of the fish. The object of the grayscale image was then segmented to separate the important and unimportant part of the object. Lastly, feature extraction process of the fish body from the calibration average value was carried out, and the estimated value of length, width, and height of fish are obtained automatically. The results show that the automatic measurement technique of fish body based on visual capture was able to produce the truth degree of accuracy of 80% to 95%.

Temu Kembali Citra Busana Muslimah Berdasarkan Bentuk Menggunakan Curvature Scale Space (CSS) Hayatun Maghfirah; Fitri Arnia; Khairul Munadi
Jurnal Nasional Teknik Elektro dan Teknologi Informasi Vol 6 No 1: Februari 2017
Publisher : Departemen Teknik Elektro dan Teknologi Informasi, Fakultas Teknik, Universitas Gadjah Mada

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (1361.781 KB)

Nowadays in Indonesia, Islamic woman's clothing has been popular and follows the latest trend. Clothing with various color, texture, and shape are available. Furthermore, the online clothing trading system is becoming more attractive, which facilitates the users the apparel images through the websites. These images can be retrieved by querying a text to the retrieval system. However, the users face difficulties in describing the clothes precisely. Thus, a retrieval method based on content, which is known as content-based image retrieval (CBIR), is developed. Here, the content is represented by color, texture, and shape. This paper aims to present and discuss an application of Curvature Scale Space (CSS) as a shape feature for Islamic woman's clothing retrieval system. The performance of retrieval results of three clothing categories is analyzed: blouse-pant, long dress, and tunic, and used different feature length. The simulations run with as many as 300 images from the three categories, 100 images from each. Performance is measured in recall and precision. The results are compared by applying another shape feature; that is the histogram of gradient (HOG). The blouse-pant group achieves the highest performance, followed by tunic and long dress categories. The different feature length affects the retrieval performance; the longer the features, the lower recall and precision values. The feature of length 4 achieves the highest performance. The CSS is applied as the feature in CBIR of Islamic clothing, results in higher performance than the HOG.

Fine Tuning CNN Pre-trained Model Based on Thermal Imaging for Obesity Early Detection Hendrik Leo; Fitri Arnia; Khairul Munadi
Jurnal Rekayasa Elektrika Vol 18, No 1 (2022)
Publisher : Universitas Syiah Kuala

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.17529/jre.v18i1.25100

Obesity is a complex disease that causes serious impact health, such as diabetes mellitus, cardiovascular disease, cancer, and stroke. An early obesity diagnosis/ detection method is required to prevent the increasing number of obese people. This study aims to: (i) fine-tune the pre-trained Convolutional Neural Network (CNN) models to build an early detection of obesity and (ii) evaluate the model performance in terms of classifying performance, computation speed, and learning performance. The thermal images acquisition procedure was conducted with 18 normal subjects and 15 obese subjects to build a thermal images dataset of obesity. Pre-trained CNN models: VGG19, MobileNet, ResNet152V, and DenseNet201 were modified and trained using the acquired dataset as the input. The training results show that the DenseNet201 model outperformed other models regarding classifying accuracy: 83.33 % and learning performances. At the same time, the MobileNet model outperformed other models in terms of computation speed with training elapsed time: 12 seconds/epoch. The proposed DenseNet201 model was suitable for implementation as an early screening system of obesity for health workers or physicians. Meanwhile, the proposed MobileNet model was suitable for mobile applications' early detection/diagnosis of obesity.

Identifikasi Tingkat Kematangan Kelapa Sawit Berbasis Pencitraan Termal Khusnul Azima; Khairul Munadi; Fitri Arnia; Maulisa Oktiana
Jurnal Rekayasa Elektrika Vol 15, No 1 (2019)
Publisher : Universitas Syiah Kuala

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.17529/jre.v15i1.12963

Indonesia is the biggest producer of palm oil (Elaeis guineenis jacq). The palm tree is a primary commodity that posses a high economic value. Palm oil must be considered in terms of quality to produce optimal and high-quality oil. Previously, the stipulation of the palm tree characterization used manual and visual image utilization method; it may have weaknesses due to the dependency of individual sorting and coruscation factor. Therefore, this research is aimed to improve the performance of the previous method in identifying the ripeness of palm tree based on thermal imaging. The excess of thermal imaging was not related to the coruscation since the level of ripeness was both determined by the temperature and colour. The detection method of this research deployed the colour-based features that are Dominant Colour Descriptor and Color Moment. The DCD and Color Moment was the input to the K-Nearest Neighbor (KNN) method. The percentage of identification rate was 89%, and the identification of oil palm maturity level using thermal imaging is more efficient because it is done without human intervention and does not depend on lighting assistance compared to manual method and method of using RGB visual images.

Title

Found 7 Documents
Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title Search

Found 7 Documents Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title

Found 7 Documents
Search