Högskolan i Skövde

his.sePublications
Change search
Link to record
Permanent link

Direct link
Kusetogullari, HuseyinORCID iD iconorcid.org/0000-0001-5762-6678
Publications (3 of 3) Show all publications
Kusetogullari, H., Yavariabdi, A., Hall, J. & Lavesson, N. (2021). DIGITNET: A Deep Handwritten Digit Detection and Recognition Method Using a New Historical Handwritten Digit Dataset. Big Data Research, 23, Article ID 100182.
Open this publication in new window or tab >>DIGITNET: A Deep Handwritten Digit Detection and Recognition Method Using a New Historical Handwritten Digit Dataset
2021 (English)In: Big Data Research, ISSN 2214-5796, E-ISSN 2214-580X, Vol. 23, article id 100182Article in journal (Refereed) Published
Abstract [en]

This paper introduces a novel deep learning architecture, named DIGITNET, and a large-scale digit dataset, named DIDA, to detect and recognize handwritten digits in historical document images written in the nineteen century. To generate the DIDA dataset, digit images are collected from 100,000 Swedish handwritten historical document images, which were written by different priests with different handwriting styles. This dataset contains three sub-datasets including single digit, large-scale bounding box annotated multi-digit, and digit string with 250,000, 25,000, and 200,000 samples in Red-Green-Blue (RGB) color spaces, respectively. Moreover, DIDA is used to train the DIGITNET network, which consists of two deep learning architectures, called DIGITNET-dect and DIGITNET-rec, respectively, to isolate digits and recognize digit strings in historical handwritten documents. In DIGITNET-dect architecture, to extract features from digits, three residual units where each residual unit has three convolution neural network structures are used and then a detection strategy based on You Look Only Once (YOLO) algorithm is employed to detect handwritten digits at two different scales. In DIGITNET-rec, the detected isolated digits are passed through 3 different designed Convolutional Neural Network (CNN) architectures and then the classification results of three different CNNs are combined using a voting scheme to recognize digit strings. The proposed model is also trained with various existing handwritten digit datasets and then validated over historical handwritten digit strings. The experimental results show that the proposed architecture trained with DIDA (publicly available from: https://didadataset.github.io/DIDA/) outperforms the state-of-the-art methods. 

Place, publisher, year, edition, pages
Elsevier, 2021
Keywords
DIDA handwritten digit dataset, Digit string recognition, Ensemble deep learning, Handwritten digit detection, Historical handwritten documents
National Category
Computer Vision and Robotics (Autonomous Systems) Computer Sciences
Research subject
Skövde Artificial Intelligence Lab (SAIL)
Identifiers
urn:nbn:se:his:diva-19396 (URN)10.1016/j.bdr.2020.100182 (DOI)000609166100006 ()2-s2.0-85098972737 (Scopus ID)
Note

CC BY 4.0

Available from: 2021-01-14 Created: 2021-01-14 Last updated: 2022-12-16Bibliographically approved
Yavariabdi, A., Kusetogullari, H., Celik, T. & Cicek, H. (2021). FastUAV-NET: A multi-UAV detection algorithm for embedded platforms. Electronics, 10(6), 1-19, Article ID 724.
Open this publication in new window or tab >>FastUAV-NET: A multi-UAV detection algorithm for embedded platforms
2021 (English)In: Electronics, E-ISSN 2079-9292, Vol. 10, no 6, p. 1-19, article id 724Article in journal (Refereed) Published
Abstract [en]

In this paper, a real-time deep learning-based framework for detecting and tracking Unmanned Aerial Vehicles (UAVs) in video streams captured by a fixed-wing UAV is proposed. The proposed framework consists of two steps, namely intra-frame multi-UAV detection and the inter-frame multi-UAV tracking. In the detection step, a new multi-scale UAV detection Convolutional Neural Network (CNN) architecture based on a shallow version of You Only Look Once version 3 (YOLOv3-tiny) widened by Inception blocks is designed to extract local and global features from input video streams. Here, the widened multi-UAV detection network architecture is termed as FastUAV-NET and aims to improve UAV detection accuracy while preserving computing time of one-step deep detection algorithms in the context of UAV-UAV tracking. To detect UAVs, the FastUAV-NET architecture uses five inception units and adopts a feature pyramid network to detect UAVs. To obtain a high frame rate, the proposed method is applied to every nth frame and then the detected UAVs are tracked in intermediate frames using scalable Kernel Correlation Filter algorithm. The results on the generated UAV-UAV dataset illustrate that the proposed framework obtains 0.7916 average precision with 29 FPS performance on Jetson-TX2. The results imply that the widening of CNN network is a much more effective way than increasing the depth of CNN and leading to a good trade-off between accurate detection and real-time performance. The FastUAV-NET model will be publicly available to the research community to further advance multi-UAV-UAV detection algorithms. 

Place, publisher, year, edition, pages
MDPI, 2021
Keywords
CNN, Deep learning, Detection and tracking, UAVs pursuit-evasion, Unmanned Aerial Vehicle
National Category
Control Engineering Communication Systems Computer Sciences
Research subject
Skövde Artificial Intelligence Lab (SAIL)
Identifiers
urn:nbn:se:his:diva-19547 (URN)10.3390/electronics10060724 (DOI)000634372800001 ()2-s2.0-85102677550 (Scopus ID)
Note

CC BY 4.0

© 2021 by the authors. Licensee MDPI, Basel, Switzerland.

Available from: 2021-03-25 Created: 2021-03-25 Last updated: 2021-04-26Bibliographically approved
Karabulut, M., Kusetogullari, H. & Kivrak, S. (2020). Outdoor Performance Assessment of New and Old Photovoltaic Panel Technologies Using a Designed Multi-Photovoltaic Panel Power Measurement System. International Journal of Photoenergy (Online), 2020, 1-18, Article ID 8866412.
Open this publication in new window or tab >>Outdoor Performance Assessment of New and Old Photovoltaic Panel Technologies Using a Designed Multi-Photovoltaic Panel Power Measurement System
2020 (English)In: International Journal of Photoenergy (Online), ISSN 1110-662X, E-ISSN 1687-529X, Vol. 2020, p. 1-18, article id 8866412Article in journal (Refereed) Published
Abstract [en]

This paper presents a new multi-photovoltaic panel measurement and analysis system (PPMAS) developed for measurement of atmospheric parameters and generated power of photovoltaic (PV) panels. Designed system presented with an experimental study evaluates performance of four new and four 5-year-old PV panel technologies which are based on polycrystalline (Poly), monocrystalline (Mono), copper indium selenide (CIS), and cadmium telluride (CdTe) in real time, under same atmospheric conditions. The PPMAS system with the PV panels is installed in Yildirim Beyazit University, Ankara Province, in Turkey. The designed PPMAS consists of three different subsystems which are (1) photovoltaic panel measurement subsystem (PPMS), (2) meteorology measurement subsystem (MMS), and (3) data acquisition subsystem (DAS). PPMS is used to measure the power generation for PV panels. MMS involves different types of sensors, and it is designed to determine atmospheric conditions including wind speed, wind direction, outdoor temperature, humidity, ambient light, and panel temperatures. The measured values by PPMS and MMS are stored in a database using DAS subsystem. In order to improve the measurement accuracy, PPMS and MMS are calibrated. This study also focuses on outdoor testing performances of four new and four 5-year-old PV panels. Average monthly panel efficiencies are estimated as 8.46%, 8.11%, 5.65%, and 3.88% for new Mono, new Poly, new CIS, and new CdTe PV panels, respectively. Moreover, average monthly panel efficiencies of old panels are calculated as 8.22%, 7.85%, 5.35%, and 3.63% in the same order. Test results obtained from the experimental system are also statistically examined and discussed to analyze the performance of PV panels in terms of monthly panel efficiencie

Place, publisher, year, edition, pages
Hindawi Publishing Corporation, 2020
National Category
Energy Engineering Energy Systems
Research subject
Skövde Artificial Intelligence Lab (SAIL)
Identifiers
urn:nbn:se:his:diva-19184 (URN)10.1155/2020/8866412 (DOI)000576103600001 ()2-s2.0-85092203591 (Scopus ID)
Note

CC BY 4.0

Available from: 2020-10-15 Created: 2020-10-15 Last updated: 2022-12-28Bibliographically approved
Organisations
Identifiers
ORCID iD: ORCID iD iconorcid.org/0000-0001-5762-6678

Search in DiVA

Show all publications