LAB: CNN Object Detection 1
Last updated
Last updated
Vehicle counting using a CNN based object detection model
In this lab, you are required to create a simple program that (1) counts the number of vehicles in the parking lot and (2) display the number of available parking space.
For the given dataset, the maximum available parking space is 13. If the current number of vehicles is more than 13, then, the available space should display as ‘0’.
The whole code should be programmed using OpenCV-Python and Pytorch.
DO NOT copy a project from online sites.
You can refer to any online material and github repository for assistance and getting ideas with proper reference citation.
Use pretrained YOLO v8.
You can also use any pretrained object detection model, such as YOLO v5
You can also train the model using custom/other open datasets
You can clone a github repository of the object detection model(e.g. YOLOv8), as long as you cite it in the reference.
Warning!
Your lab will not be scored if
If copied from the lab of previous years or from your classmates
or any other plagiarism
Download the test video file: click here to download
Need to count the number of vehicles in the parking lot for each frame
DO NOT COUNT the vehicles outside the parking spaces
Consider the vehicle is outside the parking area if the car's center is outside the parking space
Make sure you do not count duplicates of the same vehicle
It should accurately display the current number of vehicle and available parking spaces
Save the vehicle counting results in 'counting_result.txt' file.
When your program is executed, the 'counting_result.txt' file should be created automatically for a given input video.
Each line in text file('counting_result.txt') should be the pair of frame# and number of detected car.
Frame number should start from 0.
ex) 0, 12 1, 12 ...
In the report, you must evaluate the model performance with numbers (accuracy etc)
Answer File for Frame 0 to Frame 1500 are provided: download file
Your program will be scored depending on the accuracy of the vehicle numbers
TA will check the Frame 0 to the last frame
This lab will be scored depending on the Contents, Complexity, and Completeness .
You are required to write a concise report and submit the program files and the demo video.
The lab report must be written as a 'Tutorial' format to explain the whole process A to Z in detail.
Use the report template given here: https://ykkim.gitbook.io/dlip/dlip-project/report-template
Write the report in markdown ‘*.md’ format
You can also write in ''*.ipynb' format
You need to include concise explanations and codes for each process in the report
You should embed code snippets where necessary
You can also embed your demo video in the report
You must create a demo video that shows the bounding box of the cars within the parking space only.
You can submit video file to TA's email or send the download link
Zip file of report and codes
Zip file named as : DLIP_LAB_PARKING_21700000_홍길동.zip
The Zip file includes
Report (*.md) or ( * .ipynb)
Report (*.pdf)
src (source codes under /src folder)
counting_result.txt
Demo Video
Video file named as : DLIP_LAB_PARKING_VIDEO_21700000_홍길동