Traffic-Sign Recognition Project: Basis for Self-Driving Vehicle Sensing System

This is a project of cs6476 computer vision I am taking. The project is based on the open-cv library and focuses on stationary scenarios with/without noises. It is only allowed to use geometry and basic Hough methods such as HoughCircles, HoughLines, HoughLinesP to recognize the followings: traffic lights and their status, “red”, “yellow”, or “green”; stop signs; yield signs; no-entry signs; warning signs; construction signs.

Details of implementation and code are not allowed to share according to student integrity rules of the course. Here only displayed the results under artificial and real images.

Detected Green Traffic Light (Artificial Clear Image)
Detected Traffic Signs (Artificial Clear Image)
More Detected Traffic Signs (Artificial Clear Image)
Detected Traffic Signs (Artificial Noisy Image)
Detected Stop Sign (Real Image)
Detected Yield Sign (Real Image)
Detected Warning and Yield Signs (Real Image Stitched)

Comments

  1. Online Safety TrainingFebruary 21, 2022 at 9:47 AM

    You wrote this post very carefully.Traffic Control Training Online The amount of information is stunning and also a gainful article for us. Keep sharing this kind of articles, Thank you.

    ReplyDelete
  2. FastCourse4LessMarch 8, 2022 at 9:11 AM

    A very delightful article that you have shared here. Your blog is a valuable and engaging article for us, and also I will share it with my companions who need this info,California DMV Licensed Online Traffic School Thankful to you for sharing an article like this.

    ReplyDelete
  3. aaronnssdApril 1, 2022 at 4:10 PM

    You wrote this post very carefully.The amount of information is stunning and also a gainful article for us, cave diving mexico Keep sharing this kind of articles, Thank you.

    ReplyDelete
  4. DrivingcheckMay 7, 2022 at 2:17 PM

    Great job for publishing such a nice article. Your article isn’t only useful but it is additionally really informative. Thank you because you have been willing to share information with us.Cancel My Road Test Appointment Online

    ReplyDelete

Post a Comment

Popular posts from this blog

Lab Software Development 3 - Scanning Probe Microscope Controller

Image
Motivation     For home-build scanning probe systems in labs, SpecsGroup's Nanonis system is often considered a good solution with various working modes for piezo stage control. However, its' price may not be very friendly for some SPM systems, which do not need a sophisticated feedback PID loop. It is overkill especially for the non-feedback laser microwave impedance microscopy system in our lab. We had a LabView program developed by former students, but it lacks the rotational degree of freedom and line scan mode. Also, the UI's inconvenience has always been painful for users. In order to add more flexibility, I rebuilt a new control system for the SPM with QtCreator with Python as the backend. Functionalities     For convenience, all functionalities are embedded on the same page. Multithreading was implemented for scanning, imaging, and data saving. Many of the functionalities are following the workflow of Park AFM system.      Scan: By clicking "Scan", the con
Read more

Digit sequence detection of SVHN dataset

Image
  Unlike single digit detection problem like MNIST, images in SVHN dataset have more than one digit in series. SVHN Samples. Ref: http://ufldl.stanford.edu/housenumbers/ In this project, we used MSER detection and modified pre-trained VGG16 convolutional neural network to predict the digit sequence given a random image in SVHN dataset. Besides VGG16, we also used non-pre-trained VGG16 and self-built a much smaller CNN as comparisons. VGG16 Architecture. Ref: https://neurohive.io/en/popular-networks/vgg16/  Self-built CNN  The pipeline of the prediction is as: MSER detect blob regions →  CNN  find ROIs(regions of interest) → Combine overlapped/near ROIs →  CNN  Find Digits For training purposes, we mark each image with five integers as the label. The first four integers represent digits in the image, with 0 as no digit for that position and 1-10 representing 1-9 and 0. The fifth integer represents whether the area has any digit, with 0 represent a non-digit area and 1 as a window with d
Read more

Lab Software Development 2 - free-carrier DIFFUSION analysis system

Image
Quick screenshot: Paper using the software: https://www.nature.com/articles/s41467-021-25311-1 Motivation: This is a much more challenging project than the free-carrier lifetime analysis system even though this one does not include hardware control or communication. The free carrier diffusion measurement was already established and is running well. However, it is always very troublesome to do the diffusion length extractions, as our tools require three major fittings - signal response curve fitting, laser profile fitting, and convoluted Gaussian diffusion fitting. It is not possible to find a convenient commercial tool in the market to do the analysis, and it was very time-consuming to do manual analysis with Origin, Python, and other tools. A GUI would be much more convenient to use and can without a double boost the working efficiency of the lab. Environment: Since there is no control system needed, there is no need to use the lab-oriented tool LabView. Instead, I chose QtCreator and
Read more