Videoda nesne takibi için hibrit metot geliştirmesi

Loading...
Thumbnail Image

Date

2019

Journal Title

Journal ISSN

Volume Title

Publisher

Open Access Color

OpenAIRE Downloads

OpenAIRE Views

Research Projects

Organizational Units

Organizational Unit
Software Engineering
(2005)
Department of Software Engineering was founded in 2005 as the first department in Ankara in Software Engineering. The recent developments in current technologies such as Artificial Intelligence, Machine Learning, Big Data, and Blockchains, have placed Software Engineering among the top professions of today, and the future. The academic and research activities in the department are pursued with qualified faculty at Undergraduate, Graduate and Doctorate Degree levels. Our University is one of the two universities offering a Doctorate-level program in this field. In addition to focusing on the basic phases of software (analysis, design, development, testing) and relevant methodologies in detail, our department offers education in various areas of expertise, such as Object-oriented Analysis and Design, Human-Computer Interaction, Software Quality Assurance, Software Requirement Engineering, Software Design and Architecture, Software Project Management, Software Testing and Model-Driven Software Development. The curriculum of our Department is catered to graduate individuals who are prepared to take part in any phase of software development of large-scale software in line with the requirements of the software sector. Department of Software Engineering is accredited by MÜDEK (Association for Evaluation and Accreditation of Engineering Programs) until September 30th, 2021, and has been granted the EUR-ACE label that is valid in Europe. This label provides our graduates with a vital head-start to be admitted to graduate-level programs, and into working environments in European Union countries. The Big Data and Cloud Computing Laboratory, as well as MobiLab where mobile applications are developed, SimLAB, the simulation laboratory for Medical Computing, and software education laboratories of the department are equipped with various software tools and hardware to enable our students to use state-of-the-art software technologies. Our graduates are employed in software and R&D companies (Technoparks), national/international institutions developing or utilizing software technologies (such as banks, healthcare institutions, the Information Technologies departments of private and public institutions, telecommunication companies, TÜİK, SPK, BDDK, EPDK, RK, or universities), and research institutions such TÜBİTAK.

Journal Issue

Abstract

Videodaki nesnenin algılanması ve takibi, bilgisayarla görü ve görüntü işlemede önemli bir araştırma alanı olarak ortaya çıkmıştır. Nesne takibi için birçok algoritma geliştirilmiştir ve her algoritmanın başarılı veya başarısız olduğu bazı koşullar vardır. Bu tezde, videoda nesne takibi amacıyla üç nesne tespiti ve takibi algoritmasından oluşan güçlü bir karma sistem önerilmiştir. Bunlar şablon eşleştirme, renk histogramı ve özellik çıkarımına dayalı SURF algoritmalarıdır. Bu algoritmaları hibrit sistemde uygulamak için OpenCV kütüphanesi kullanılmıştır. Algoritmalar uygulanırken; gaussian blur, renk uzayı dönüşümleri, Otsu eşiklemesi, kayan pencere yaklaşımı, özellik çıkarımı ve betimlemesi, ve uzaklık hesaplamaları gibi farklı teknikler uygulanmıştır. Videodaki herhangi bir nesne seçilebilir ve seçilen nesne videonun geri kalanında takip edilebilir. Nesnenin tıkanmasını önlemek ve sahnenin ani hareketinin etkilerini en aza indirmek için, videonun her beşinci karesinde seçilen nesnenin yenilenmesi yaklaşımı kullanılır. Hibrit sistemin amacı, video karelerindeki takip edilecek nesnenin tespit oranını iyileştirmektir. Tüm performans testleri NTU-VOI 2018, Visual Tracker Benchmark 2013, NfS 2017 ve Davis 2017 veri setleri üzerinde gerçekleştirilmiştir. Önerilen hibrit sistemin test sonuçları, üç ayrı tespit ve takip algoritmasının sonuçlarıyla karşılaştırılmıştır. Sonuçlar, hibrit sistemin video nesne takibi için işlem süresi dışında en iyi performansı verdiğini göstermektedir.
Detecting the object in the video and tracking it has been emerging as an important research field in computer vision and image processing. Many algorithms have been developed for object tracking and there are some conditions in which each algorithm is successful or unsuccessful. In this thesis, a robust hybrid system that consisting of three object detection and tracking algorithms is proposed for the purpose of tracking object in video. These algorithms are template matching, color-based histogram and SURF based on feature point. OpenCV library have been used to implement these algorithms in hybrid system. While implementing algorithms, different techniques have been applied such as gaussian blur, color space conversions, Otsu thresholding, sliding window approach, feature extraction and description, and distance measurements. Any object from the video can be selected and the selected object can be traced in the rest of the video. To prevent occlusion of the object and to minimize the effects of sudden movement of scene, refreshing selected object approach is used each fifth frame of the video. Aim of the hybrid system is to improve the detection rate of the object to be tracked in sequence of video frames. All performance tests have been performed on NTU-VOI 2018, Visual Tracker Benchmark 2013, NfS 2017 and Davis 2017 datasets. The test results of the proposed hybrid system have been compared with the results of the three individual detecting and tracking algorithms. The results show that hybrid system gives the best performance except for processing time for tracking object in video.

Description

Keywords

Bilgisayar Mühendisliği Bilimleri-Bilgisayar ve Kontrol, Nesne izleme, Computer Engineering and Computer Science and Control, Renk histogramları, Object tracking, Color histograms, Şablon eşleme, Template matching

Turkish CoHE Thesis Center URL

Fields of Science

Citation

WoS Q

Scopus Q

Source

Volume

Issue

Start Page

0

End Page

102