Search or add a thesis

Advanced Search (Beta)
Home > Road Obstacle Detection

Road Obstacle Detection

Thesis Info

Author

Safian Haq, Usman Ahmad

Program

BS

Institute

Habib University

Institute Type

Private

City

Karachi

Province

Sindh

Country

Pakistan

Thesis Completing Year

2019

Thesis Completion Status

Completed

Subject

Engineering

Language

English

Added

2021-02-17 19:49:13

Modified

2024-03-24 20:25:49

ARI ID

1676724381295

Asian Research Index Whatsapp Chanel
Asian Research Index Whatsapp Chanel

Join our Whatsapp Channel to get regular updates.

Similar


Autonomous vehicle systems can be divided into two main parts, the perception system and the decision-making system. In this project, the goal was to focus on the perception system’s subsystem which was an object detector for road obstacle detection on the roads of Karachi. To do this, firstly a dataset of 3000 images was annotated with bounding box annotation for 12 different kinds of road obstacles. These images were extracted frames (every 10th second) from about 10 hours of dashcam footage from different areas of Karachi. In parallel three different models were trained on the Berkeley Deep Drive Dataset (BDD100k), which were YOLOv3, RetinaNet and Faster R-CNN. Due to computational constraints the models were trained on only 5,000 out of 70,000 images and validated on 1,000 out of 10,000 images present in the BDD100k dataset. The models trained had the following mAP on BDD100k; YOLOv3(29.47), RetinaNet(37.34) and Faster R-CNN(35.78). These models were then used as pretrained models for transfer learning on Karachi Dataset to create three new models. The models trained on this new dataset had the following mAP; YOLOv3(41.67), Retina Net(67.26), Faster R-CNN(65.80). Analysis suggests that transfer learning using the BDD100k dataset, is the most optimum technique for training an object detection model on Karachi Dataset
Loading...
Loading...

Similar Books

Loading...

Similar Chapters

Loading...

Similar News

Loading...

Similar Articles

Loading...

Similar Article Headings

Loading...

سلام

سلام
جس نے ہجر کی خندق سے!
وصال کی نقیب ناقہ کھینچ کر لاتے ہوئے!
آپ چاندنی جیسی سفارت سے!
سوالی اسرار اور گرد آلود بھید کے منہ دھوتے ہوئے!
بے یقین دست کی رگوں میں!
سنہری تبسم۔۔اثبات وفا کی زندگی کو بھر دیا
اس ’’الحسین منی‘‘ کی حقیقت پہ لاکھوں سلام
جس نے مزاج سبز بہار کی کتاب سے
چادر تطہیر سے گرد صاف کی
جس نے درِ علم سے علم کے شہرتک!
طاق ساعتوں کیساتھ پاسبانی کرتے ہوئے!
قاتلین، منافقین اور قابضین کے چہرے دکھا دیے
جو آج بھی لا مکاں۔۔۔عصرے رواں میں!
یزیدیت کی دھجیاں اْڑا رہا ہے

اس قتیل نینوا پہ لاکھوں سلام
جس نے بنجر وادیوں کی طرف!
سبز خوشبو کا رخ موڑ دیا
جس نے فاصلوں۔۔۔ساعتوں کو انگلیوں پر نچاتے ہوئے!
ویران دشت کا رشتہ!
الوہی سبزہ گاہ سے جوڑ دیا
جس نے لسان فلک کے لہجے میں!
’’و اَنا مِن الحسین‘‘ کی تشریح نوک نیزہ پہ کی
اس حسین ؑابن حیدر پہ لاکھوں سلام
بنتِ حسینؑ و علیؑ پہ لاکھوں سلام

Impacts of Psychological and Domestic Violence On Women in Pakistan: Problems & Solutions in the Light of Islamic Teachings

Since the creation of woman, she faces many problems in her life. Different societies have their own customs and traditions. And woman faces problems regarding them. Pakistani society has its own influence and civilization which causes many problems of women. In these traditions, one of the bad behaviors is, marriage of woman on wrong time i.e. Late marriage or early time marriage. In the result, at least, she faces Problems regarding dowry, Joint family system, Family disintegration, Childlessness, Propensity to violence, Effects of husband remaining alone from wife etc. On the basis of social divisions in Pakistani family system and depiction of woman issues having effects on herself, the significant and their mediation is very necessary, too. Many of these problems has Psychological impacts on woman in her domestic life. In Pakistani society where woman faces domestic and family problems, there economic problems too pester her which include greed for riches and lack of them both pester her psychologically. In this paper, above mentioned problems of women in Pakistani society has been discussed in the light of Islamic teachings.

Meshfree Methods for the Numerical Solutions of Partial Differential Equations

In this dissertation, meshfree (meshless) methods using meshless shape functions are proposed for the numerical solutions of partial differential equations (PDEs). These PDEs have either integer or fractional order time derivatives. Weighted θ-scheme (0≤θ ≤1) is used for time discretization of integer case, whereas, for fractional case, the same discretization scheme is combined with a simple quadrature formula. For space (spatial) discretization we used meshless shape functions owing Kronecker delta function property. These shape functions are obtained viapointinterpolationapproachandradialbasisfunctions(RBFs). Finallywiththehelpofcollocationmethodthe given PDE reduces to system of algebraic equations, which are then solved via LU decomposition in iterations. For the proposed numerical scheme, stability analysis is carried out theoretically and computational examples are provided to support the analysis. The proposed scheme has been tested via application to several concrete and benchmark problems of engineering interest. ApproximationqualityandaccuracyofcomputedsolutionsaremeasuredusingL∞, L2 andLrms discrete errornorms. Efficiencyandorderofapproximationoftheproposedschemeinspaceandtimeareanalyzedthrough variation of number of nodal points N and time step-size δt. The documented results, in the form of tables and figures, reveal very good agreement to true solutions as well high accuracy to earlier proposed technique available in the literature. In RBFs, the presence of shape (support) parameter c∗ plays a crucial role. Accuracy of the RBFs based scheme can be improved via proper selection of this parameter. For this purpose, an automatic optimal shape parameter selection algorithm is proposed. To check effectiveness and automatic (adaptive) nature of this algorithm in RBFs approximation method, time fractional Black-Scholes models have been solved. It has been noted that the proposed algorithm worked well and gives excellent accurate solutions for various fractional order time derivatives. The RBFs approximation (Kansa) method results in dense ill-conditioned matrix. For the treatment of this issue weproposeahybridRBFs(HRBFs)approximationmethod. Byextendingthisidea, anadaptive(automatic)algorithm is proposed for optimal parameters selection in HRBFs. For validation, again time fractional Black-Scholes models are reconsidered. Simulations revealed acceptable accurate solutions in hybrid RBFs method too. Along with that significant reduction in condition number of the resultant matrix is observed up to several manifold. Hence, HRBFs method can be seen as an alternative remedy for curing ill-conditioning in usual RBFs method. Computer simulations have been carried out via MATLAB R2013a on a personal laptop with configuration, Processor: Intel(R) Core(TM) i5-5200U CPU @ 2.20GHz 2.20GHz, RAM: 4.00 GB, System type: 64-bit Operating System, x64-based processor.