2. Q1: How precise image recognition is done?
・It use sensor technology to keep track of its surroundings.
・It is capable of making all kinds of decisions related to driving.
・It can be driven day or night, even on complex roads.
3. Q2:What is the basis for the development of driverless cars?
・FPGA(Field Programmable Gate Array) is logic circuits programmed by designers
in the field.
・PL(Programmable Logic) is hardware logic to implement image recognition for
human detection.
・PS(Processor System) is image processing software for lane keeping, navigation,
and motor control.
4. Q3: How to control the space between cars?
・CACC(Cooperative Adaptive Cruise Control) is a module
of the control space between cars.
・We can reduce the traffic jam to use of technology of
driverless car.
・We think that “Can it solve the problem of change lane”.
5. Article 1 #1 Main Point
・Creating a model that exceeds human drive capacity would save many lives.
・Clear white lines on the road make it easier to recognize.
・Randomly change the brightness to simulate various light conditions and have
the model learn from them.
6. Article 1 #2 Related the Ted Talk
・Grasp and anticipate what is going on around you.
・It can be driven on highways and both day and night.
・Cameras and sensors are used to obtain information about the surroundings.
7. Article 1 #3 Answer our research question
・The steering angle operated by a human is acquired and learned from the camera
image.
・They simulate various situations and learn to deal with many situations.
・Creating a model that exceeds human drive ability would eliminate the loss of
many lives.
・The system can be operated automatically regardless of the season.
8. Article2 #1 Main Point
・Two FPGAs are required for system implementation.
・One is called Arty Z7 and feeds the webcam images.
・The second one is called Zybo Z7-20 and it returns the recognition results of the
BNN hardware.
9. Article2 #2 Types of Image Processing
・Lanes, intersections, obstacles, traffic lights, humans, etc.
・Each of these must be processed after image recognition.
10. Article2 #3 Answer our research question
・Further development of both hardware and software will further increase the
accuracy of image recognition.
・Increasing the accuracy of image recognition is the best way to make fast
decisions in case of unforeseen situations.
・
11. Article3 #1 Main point
・This report talked about “How to control to change the lane anytimes”
・CACC has a problem that it use only load which do not have traffic jam.
・MSCACCLC can makes space to change lane anytime.
・Also it can reduce the distance between cars.
12. Article3 #2 Related the Ted Talk
・In Ted Talk, if machine control the car, we can reduce the traffic jam.
・If machine reduce the car distance, we can reduce the commuting time.
・MSCACCLC can reduce the not only commuting time, but also reduce the traffic
jam.
13. ・Now there are technology about changing the load with automatic.
・That technology can reduce not only car accident but also reduce traffic jam.
・This problem is trivial but we need think about this problem once.
Article 3 #3 Answer our research question
14. Discussion and Implications of our findings#1
・The image recognition system can acquire and learn the steering angle operated by
humans from camera images.
・The key to improving the accuracy of image recognition technology is further
development of hardware and software.
・Technology that automatically maintains a distance between vehicles using sensor.
15. Discussion and Implications of our findings#2
・It will be safer and easier to move than ever before.
・It reduces the stress of driving and reduces the burden on health.
・Avoid traffic jams and move more efficiently.
16. Conclusion
・Mainly questioned the safety of driverless cars and went into the investigation
・Deepened our understanding of the answers to questions and their content
・Considered what benefits driverless cars would bring to mankind
17. References
Articles
[1]Du, Shuyang, Haoli Guo, and Andrew Simpson. "Self-driving car steering angle prediction based on image recognition." arXiv preprint
arXiv:1912.05440 (2019).
[2]Kojima, Akira, and Yohei Nose. "Development of an autonomous driving robot car using FPGA." 2018 International Conference on
Field-Programmable Technology (FPT). IEEE, 2018.
[3]WANG, Haoran, et al. Make space to change lane: A cooperative adaptive cruise control lane change controller. Transportation Research
Part C: Emerging Technologies, 2022, 143: 103847.
Video
Google's driverless car
URL: https://www.ted.com/talks/sebastian_thrun_google_s_driverless_car