Multiple Object Tracking Based On Firefly Algorithm And FSIM
Publish place: 1st National Conference on Development of Civil Engineering, Architecure,Electricity and Mechanical in Iran
Publish Year: 1393
نوع سند: مقاله کنفرانسی
زبان: English
View: 1,429
This Paper With 5 Page And PDF Format Ready To Download
- Certificate
- من نویسنده این مقاله هستم
استخراج به نرم افزارهای پژوهشی:
شناسه ملی سند علمی:
DCEAEM01_175
تاریخ نمایه سازی: 18 دی 1393
Abstract:
Multiple object tracking is an active area of research in computer vision. Using different strategies such as optimization solutions, estimators, filters and other solutions that can be adapted to the tracking problem, is aim for many researchers to achieve more acceptable results. This paper introduces new approach to improve accuracy of multiple object tracking based on optimization solution using firefly algorithm. Firefly is a new meta heuristic optimization algorithm. It has some characteristics like PSO algorithm with better performance in optimization problems. In this approach, the similarity between the primary objects window and the secondary objects window in each frame selected as the objective function. Then, we use FSIM algorithm to calculate the similarity between two windows. At the end, problem of locating these windows in each frame is solving by Firefly optimization algorithm. Experimental results approve that the proposed method improves greatly the tracking accuracy during different challenges, like occlusion, low quality, complex background and Unanticipated movements.
Keywords:
Authors
moslem Arababadi
Electrical and computer engineering department Semnan University Semnan, Iran
sasan azadi
Electrical and computer engineering department Semnan University Semnan, Iran
مراجع و منابع این Paper:
لیست زیر مراجع و منابع استفاده شده در این Paper را نمایش می دهد. این مراجع به صورت کاملا ماشینی و بر اساس هوش مصنوعی استخراج شده اند و لذا ممکن است دارای اشکالاتی باشند که به مرور زمان دقت استخراج این محتوا افزایش می یابد. مراجعی که مقالات مربوط به آنها در سیویلیکا نمایه شده و پیدا شده اند، به خود Paper لینک شده اند :