100 sets of computer graduation projects
WeChat Mini Program Project Practice
If you need the source code, drop me a message
Table of contents
1. Background, purpose and significance of topic selection
2. Application principles and theoretical basis
3. Plan argumentation analysis
4. Topic Features and Expected Results
4.1 Features of topic selection
1. Background, purpose and significance of topic selection
Due to the rapid development of society, people rely more on cars for travel, and the use of cars has also increased significantly. Since the 1990s, our country's economy has developed vigorously, and the level of urban modernization and road traffic motorization has significantly improved. Safe driving issues It has also gradually attracted close attention from central and local social institutions and scientific research institutions. Frequent traffic accidents have seriously affected people's lives. According to incomplete statistics, nearly 1 million people are killed every year due to traffic accidents, and countless others are injured [2]. After investigation and research, people found that there are three main factors that affect traffic safety, namely drivers, roads and vehicles. If there is a problem with the vehicle and it is not eliminated in time or the road potholes are severely damaged, it may cause a traffic accident. However, these two factors are not the most important, because in general, whether it is roads or railways, the construction technology is very mature, and there are no problems with vehicles. The road factor has little impact on traffic safety unless weather conditions are very severe. However, so far, the incidence rate of traffic accidents has remained high around the world. It can be seen that drivers account for the main factor in the occurrence of traffic accidents [3]. Poor driving skills or poor driving conditions of drivers will affect traffic safety. The short-term "micro-sleep" phenomenon is caused by driver fatigue and is an important cause of traffic accidents. Therefore, if a system can be developed to provide early warnings in the early stages of driver fatigue, it will undoubtedly greatly reduce traffic accidents. incidence rate.
At present, with the rapid development of digital image processing technology, high-performance CPU and large-capacity memory, it is possible to use digital image processing technology to automatically detect and judge fatigue status. And with the development of science and technology and the proposal of green themes, low power consumption, high adaptability, low cost, and high integration have become important development trends in electronic products, and the use of ARM, DSP, FPGA, etc. is also increasing. Although ARM has a strong ability to manage transactions, ARM has a weak ability to process a large number of signals. DSP has strong ability to process data, but the processing speed is not ideal. In contrast, FPGAs using SOPC on-chip programmable technology have the characteristics of miniaturization, low power consumption, and high speed. In addition, FPGAs have short development cycles, low investment in development software, high reliability, stable quality, and standard products do not require testing. And it has the inherent characteristics of real-time online detection and on-site programming. With the continuous development and improvement of modern technology, the cost of FPGA chips is getting lower and lower. These advantages make people use FPGAs more and more frequently. FPGAs are increasingly used in the field of large data volume and high-speed signal processing. middle.
In summary, based on today's realistic background and technical background, a driver fatigue driving detection system based on OpenCV was developed to promptly warn the driver when the driver is in a "micro-sleep" state, thereby preventing social traffic To prevent accidents, save lives, maintain social traffic safety and order, and maintain the happy life of every family.
2. Application principles and theoretical basis
Through the investigation of image recognition and related technologies, this article understands user needs, as well as image recognition and feature extraction based on OpenCV, as well as the corresponding drivers. According to the human eye fatigue algorithm detection process, software analysis, solution design, and design and Implementation, system integration development and debugging, system operation and debugging. Using OpenCV for human eye detection, through a website written using a python program, calling the camera to capture faces, some research has been done on feature detection and target extraction. Based on the use of the AdaBoost cascade algorithm based on Haar features, through The facial image collected by the camera undergoes a series of image processing such as scanning, human eye search, and human eye positioning. Within the basic range of the human face, the human eye is positioned according to the horizontal projection and the distance between the upper and lower eyelids is calculated, and then based on the given PERCLOS Standard, analyze and judge the distance between human eyelids, so as to make accurate judgments on fatigue driving. Finally, according to the PERCLOS standard, the human eye state at the specified moment is judged whether the driver is fatigued. If so, a warning message is given. Information is fed back through the python page.
3. Plan argumentation analysis
This part mainly includes the difficulties of the subject and key issues to be solved, the research methods to be adopted and their feasibility, etc.
The face detection block diagram is as follows:
The difficulty in system design and the key issue to be solved is the positioning of eyes. Everyone's face includes features such as chin, mouth, nose, eyes, eyebrows and so on. Typically, these areas have darker grayscale values than surrounding areas.
Therefore, the research methods proposed to address this issue are:
The position of the eye window is roughly positioned by performing horizontal grayscale integral projection on the upper 2/3 of the face of the original grayscale image. Due to different people, faces have minimum values at the eyebrows and eyes, but the relationship between the two minimum values at the eyebrows and eyes is not fixed.
When locating the eye window, first obtain the minimum and sub-minimum values of the horizontal grayscale integral projection of the original image, determine the rows where the minimum and sub-minimum values are located, and use the upper boundary of the line where the eyebrows are located, and the eyes and A value of 2 times the difference between the eyebrow rows is used to roughly locate the human eye. This can be predicted: after actual system testing on different people at different times, it can be concluded that this algorithm can not only reduce the computational complexity, but also effectively coarsely locate the eye window.
4. Topic Features and Expected Results
4.1 Features of topic selection
The advancement of the Internet has made all social activities accurate to the level of machine operation, and it has also played a big role in the automobile driving industry. This system can judge the driver's fatigue level through face recognition, which has largely eliminated traffic congestion. In the event of an accident, the driver's degree is instantly verified by uploading face pictures and real-time shooting.
4.2 Expected results
(1) Successfully set up the environment and debug it;
(2) Complete code that can realize specific functions;
(3) Graduation thesis.
5. Work progress
Week 1: Determine the graduation topic, complete the double selection of teacher and student graduation projects, and configure the environment to review relevant books and materials.
Weeks 2-3: Consult technical information and English literature related to the graduation topic, learn relevant algorithms, and complete the proposal report.
Week 4: Planning process, starting to prepare thesis materials.
Week 5 to the end of winter vacation: Find relevant information, and write and test the face classification and image encryption parts. Start writing your graduation thesis.
Weeks 6-7: Prepare for self-examination and writing of the paper.
Weeks 8-11: Bug hunting and debugging based on previous work.
Weeks 12-14: Test and maintain functions, analyze experimental results, and improve algorithms.
Weeks 15-18: Self-examination and inspection of work, writing and checking the paper in the later stage of the thesis, modifying the format, finalizing the paper, and preparing presentations for the defense, and preparing for the defense.
Week 19: Sort out the system design and thesis, become familiar with the defense process, and conduct thesis defense.
6. Main references
[1] Xie Xiuzhen. Research on driver fatigue detection based on machine vision D], Central South University, 2020
[2] Li Dong. Design and implementation of driver fatigue detection system based on FPGA ID], Dalian University of Technology, 2019
[3] Quantitative similarity between the cognitive psychomotor performance decrement associated with sustained wakefulness and alcohol intoxication[Al. In Hartley, L.R.(Ed.ManagingFatigue inTransportation Proceedings ofthe Third InternationalConference on Fatigue and TransportationFremantle, Western Australia[C],Oxford UK. Elsevien Science Ltd, 2019
[4] Li Wei, He Qichang, Fan Xiumin. Driver fatigue state detection based on vehicle control signals [J], Journal of Shanghai Jiao Tong University, 2020, 44(2):292-296
[5] Zhang Liwen, Yang Yanfang, et al. Fatigue driving detection based on facial features [J], Journal of Hefei University of Technology: Natural Science Edition, 2021, 36(4):448-451
[6] Weng Maorong, Li Qiang, Hua Caixia. Research status and development trends of motor vehicle driver fatigue detection systems [J], Journal of Zhejiang Industry and Trade Vocational and Technical College, 2019, 6(1):52-56
[7] Hu Shifeng. Research on real-time detection algorithm of fatigue driving based on eye characteristics [D], Shanghai Normal University, 2021
[8] Zhao Ruxin, Song Chunlin. Real-time fatigue driving warning algorithm based on Adaboost and LBP algorithm [J]. Microcomputer Applications, 2022, 38(05): 1-15
[9] Australian Transport Regional Policy Section. Fatigue management for commercialvehicle drivers[EB/OL], 2018