1.1 Course Introduction

1.1.1 Introduction to Sensor fusion and non-linear filtering

definition : we should use a sequence of noisy observations from one or more sensors.

• If we use one sensor, we call it filtering.
• And if we use more sensors, we call it sensor fusion.

goal : filter the sequence of noisy sensor observations to get a better estimate of some unknown quantity of interest(=state.)

However, we are not just content with an estimate.

We are also interested in finding the associated uncertainty measures, that is, how sure we are of our estimate.

CAMERA

• 장점 : detecting and classifying, determining the relative angle
• 단점 : cannot be measured directly using a single camera.

radar

• 장점 : good at measuring the distance to an object, Doppler shift(=relative radial velocity)
• 단점 : not good at measuring the angle to an object.

1.1.2 Demonstrations

1.1 Course Introduction pt II MED INTRO V2-en

본 과정에서 배운것의 활용 분야

• 데모 1 : 카메라 기반 위치 속도
• 데모 2 : 카메라 radar 기반, road geometry estimation
• 데모 3 : Self-localization

1.1.3 Course structure and learning outcome

1.1 Course Introduction pt III MED INTRO V2-en

강의 구성 및 간략한 설명

1.1.4 About the course staff

• LARS HAMMARSTRAND

• ERIK STENBORG