Foram Patel

 

Date:  Tuesday, September 24, 2019

Time:  10:30 am – 12:30 pm

Location: Room 3105 Lambton Tower

While an autonomous car drives by itself on the road, the types of objects surrounding the car can be dynamic (e.g., cars and pedestrians), stationary (e.g., buildings and benches), and variable (e.g., trees) depending on if the location or shape of an object changes or not. Different from the existing approaches that use image-based techniques for the detection and recognition of objects in a scene, this thesis research proposes an approach of eliminating stationary and variable objects from the scene to allow the autonomous car focusing on those dynamic objects that may cause danger to its driving. This approach uses the prior knowledge of non-dynamic objects in a virtual city and verifies their existence in a real-time image by matching feature points between the virtual and physical objects. Verified objects are then eliminated from the image with a combined algorithm using contour detection and class activation map, which is anticipated to improve the efficiency and accuracy when recognizing dynamic objects.

 

Thesis Committee: 

 

MSc Thesis Defense Announcement

 

5113 Lambton Tower, 401 Sunset Ave., Windsor ON, N9B 3P4, (519) 253-3000 Ext. 3716 csgradinfo@uwindsor.ca