Capturing the Future: How Cameras Drive Autonomous Vehicles

Hey dude...!

I Am back..😉😉🤘
I'll tell you some knowledge shear about Camera's
These things all about Self-Driving Cars 🚨🚨

I think your also interested & enthusiastic like me

Camera's are passive systems, senses environment based on energy photon's..!🤏🤏

Cameras Are one of the Main sensors used in Self-Driving Cars, They are very good at high resolution tasks. They’re relatively cheap compared to Radar and Lidar. Cameras are great for classifying objects and understanding a scene.

see this blog and clear your doubts about camera calibration

Link

Cameras work on the basic principle of landing the light emitted from the objects on a photosensitive surface through a lens. The photosensitive surface also called the image plane is where the light is stored as an image. After a light passes through the lens, the light sensitive surface converts those rays into electrons, those electrons are then converted to voltage, amplified, passed through an Analog-Digital Converter (ADC) and converted to a number in a metric unit or pixels.

The projected points are converted from the metric unit to pixels so the image can be used for further processing and extracting information LIKE detection, classification, tracking of objects that are in the path of our autonomous vehicle🚗.

Cameras are well-known sensors used to output things like bounding boxes, lane positions, traffic light colors, and road signs. Cameras are never out of the question in self-driving cars.

camera can be used for autonomous vehicle applications, each individual camera should be calibrated. This is due to the use of lens in the camera sensor which cause distortion in the images they create. These distortions are rectified by using pictures of known objects like checkerboards, since these pictures dimensions and geometry is known by comparing the camera output image with the original pictures we can compute the calibration parameters which will account for those distortions. Two types of these distortions are Barrel and Pincushion as shown below.

There are two types of camera chips on the market today: CCD and CMOS.

-- CCD is an abbreviation for Charge-Coupled Device,

-- CMOS is an abbreviation for Complementary Metal Oxide Semiconductor.

Types of cameras:

1) full hd 1080p Camera's 30..or more FPS

2) Fish Eye Cameras

3) Mipi cameras

4) 3D Camera's

5) grayscale cameras

6) depth Cameras

7) IR camera's

Advantages of Camera's:-

Cameras offer the benefits associated with high-definition images, but they don't perform well in all types of weather. Also, the camera only captures visible visual data.

- Camera's needs to good lighting conditions

-- cameras provide more data compare than lidar and radar

--- Low cost

---- easy to replace & fix issues and easy fix positions of autonomous cars

------ low power consumer than compares lidar and radar and another sensor's in Av

------ easy to update features

------- cameras are Fusion(combination) into Lidars & Cameras; Radar & Cameras in development of perceptions tasks..

----------update rate = 30~60fps

Disadvantages of Camera's:-

- it's probably night not so much light conditions, that why in night times it's can be performed poorly.

-- only Cameras is not good for human safety (my own thoughts)

---But the problem is, they don’t provide us with ‘Depth Information’ as LiDARs do:)

Why TESLA only use camera's, why not use Lidar & RADAR

Tesla uses:- see this video, you have to know something..!

TESLA NEW CARS hardware & software's:-

Tesla thinking about all modules of various sensors to replace cameras. What I mean Lidar replace to cameras & radar replace to cameras, that time cameras behave like lidar & radar work....! Is that's possible, I don't know, this thing answers in the Future, and also time decides everything.

2D, 3D, Enough for Now, in the Future (4D):- see this

A team of scientists from Stanford University has created the best "eyes" for self-driving cars and delivery drones so far. A 4D camera that can capture information in almost 140 degrees, he can collect more information in one image than a conventional camera. The researchers call their design "the first single-lens light-field camera with a wide field of view." Additional information to bring the results to 4D relies on light field photography. That means he can observe and record the direction and distance of the light hitting the lens and combine it with the resulting 2D image of him.

MUST WATCH--

LAST WORDS:-
One thing to keep in the MIND Ai and self-driving Car technologies are very vast...! Don't compare yourself to others, You can keep learning..........

Competition And Innovation Are Always happening...!
so you should get really Comfortable with change...

So keep slowly Learning step by step and implement, be motivated and persistent