Evaluation of Image Detection Techniques Acquired from Camera Images

Basit Öğe Kaydı
dc.contributor.authorErkartal, Buğra
dc.contributor.authorYılmaz, Atınç
dc.date.accessioned2026-01-31T14:59:21Z
dc.date.available2026-01-31T14:59:21Z
dc.date.issued2025
dc.departmentİstanbul Beykent Üniversitesi
dc.description.abstractIn terms of situational awareness, object recognition, and real-time decision-making, abstract camera-based image detection methods have grown to be a core element of autonomous driving systems. This study presents a comprehensive evaluation of camera-based object detection techniques used in autonomous driving systems. Traditional methods such as Haar Cascades and HOG are reviewed alongside modern deep learning architectures including CNN, YOLO, and GANs. The study examines their strengths, weaknesses, and real-time performance across various detection tasks such as 2D/3D object detection, semantic/instance segmentation, and behavioral prediction. Especially promising for improving perceptual dependability under demanding environmental conditions and sensor fusion techniques combining data from lidar, radar, and cameras. By forecasting pedestrian and vehicle movements, deep learning-based behavioral prediction systems also greatly help to enable safer and more proactive driving. The results show that application-specific needs including accuracy, computational efficiency, and real-time processing should direct the choice of the suitable object identification technique. The findings suggest that no single technique is sufficient on its own; rather, the fusion of multiple systems, supported by adaptive and resource-efficient architectures, is crucial for safe and reliable autonomous driving. The research highlights the need for modular and scalable perception solutions capable of adapting to real-world complexities. Future studies should concentrate on the creation of low-cost, adaptive, multi-modal perception systems, which are fundamental for the safe and broad implementation of autonomous driving technology.
dc.identifier.endpage43
dc.identifier.issn2548-1185
dc.identifier.issn2587-2176
dc.identifier.issue2
dc.identifier.startpage33
dc.identifier.urihttps://hdl.handle.net/20.500.12662/10301
dc.identifier.volume9
dc.language.isoen
dc.publisherNişantaşı Üniversitesi
dc.publisherNisantasi University
dc.relation.ispartofInternational Journal of Engineering Science and Application
dc.relation.ispartofInternational Journal of Engineering Science and Application
dc.relation.publicationcategoryMakale - Ulusal Hakemli Dergi - Kurum Öğretim Elemanı
dc.rightsinfo:eu-repo/semantics/openAccess
dc.snmzKA_DergiPark_20260128
dc.subjectAutonomous Vehicle Systems
dc.subjectOtonom Araç Sistemleri
dc.titleEvaluation of Image Detection Techniques Acquired from Camera Images
dc.typeArticle

Dosyalar

Koleksiyon

Öksüz Yayınlar Koleksiyonu