HARAKATDAGI VA SHOVQINLI MUHITLARDA YUZNI ANIQLASH ANIQLIGINI OSHIRISHNING GIBRID YONDASHUVI
Keywords:
yuzni aniqlash, kompyuter ko‘rish, gibrid yondashuv, CNN, Haar Cascade, sun’iy intellekt, videoanalitika, real vaqt tizimlari., face detection, computer vision, hybrid approach, CNN, Haar Cascade, artificial intelligence, video analytics, real-time systems.Abstract
Mazkur tezisda harakatdagi va shovqinli muhitlarda yuzni aniqlash aniqligini oshirish muammolari tahlil qilinadi. Zamonaviy kompyuter ko‘rish tizimlarida yuzni aniqlash algoritmlarining samaradorligi yorug‘lik o‘zgarishi, kamera tebranishi, tasvir shovqinlari hamda obyekt harakati sababli pasayadi. Shu sababli tezisda klassik kompyuter ko‘rish algoritmlari va sun’iy intellektga asoslangan chuqur o‘rganish modellarini birlashtiruvchi gibrid yondashuv taklif etiladi. Haar Cascade va CNN algoritmlarining integratsiyasi orqali real vaqt tizimlarida aniqlik va tezkorlikni oshirish imkoniyatlari ko‘rib chiqiladi.
This paper analyzes the challenges of improving face detection accuracy within dynamic and noisy environments. In modern computer vision systems, the efficiency of face detection algorithms degrades due to illumination changes, camera shake, image noise, and object motion. To address these issues, this paper proposes a hybrid approach that integrates classical computer vision algorithms with artificial intelligence-based deep learning models. By combining Haar Cascade and Convolutional Neural Networks (CNN), the potential to enhance both accuracy and processing speed in real-time systems is demonstrated.
References
