2019 - Arv2Counter, Software Tool to Detect and Count Vehicles in Urban Roads using Computer Vision
Arv2Counter was developed for the detection and counting vehicles in urban roads ArV2Conter, which uses a set of computer vision techniques for the subtraction, improvement, tracking and counting of moving objects that they are considered vehicles.
In this way, it allows the user to collect the analysis of vehicular traffic of an urban road, and then statistical reports can be made for periods of time such as time slots and days of the week.
Arv2Counter has a user interface to configure the software tool, count the vehicles and generate the statistical reports using the Python programming language and the PyQt library. Finally, field tests are carried out with an AXIS 214 PTZ IP camera to validate the performance of the tool for the effects of light reflection and of areas with saturation and shadows, generated by lighting changes in the scene.
Actually, ArV2Counter is a registered software legally recognized by the Interior Ministry of Colombia. This document can be downloaded here.
ArV2Counter GUI is divided into three modules:
1. Configuration,
2. Vehicle counting and
3. Reports.
The figure at left shows part of the configuration module where users are able of:
1. Defining the source of data:
a. Camera
b. Image sequence
c. Video
2. Calibrating the camera in order to define the intrinsic and distortion parameters.
In the same configuration module, other properties that control the behavior of ArV2Counter can be defined:
1. To select the area of interest in the image where vehicles will appear.
2. To define the parameters for background subtraction:
a. Number of images in the history
b. Number of Gaussian functions.
3. To define the tracker algorithm parameters:
a. Vehicle sense of motion.
b. Minimum blob size in pixels.
c. Maximum number of corners to track by blob.
4. To define the pre-processing parameters:
a. Erosion operator.
b. Dilatation operator.
All these parameters are stored in a MySQL database.
Once the configuration parameters of all algorithms used in ArV2Counter are defined, the vehicle counting module can be initiated.
This module was implemented in a second tab in the ArV2Counter GUI. In this tab, users are able of:
1. Panel to save the results of vehicle counting.
2. Controls to play, pause and stop the counting algorithm.
3. Displaying information about vehicles currently counted.
4. Selecting the path and filename where the resulting video will be stored.
5. Displaying the current date and time.
Then, after the vehicle counting session was performed, ArV2Counter can generate a PDF report about a particular session. To do so, the reports GUI offer a basic and an advanced interface.
In the basic report interface users are able of:
1. Selecting a range of hours to generate the report.
2. Adding user comments.
3. Observing a table with the data extracted from the database, considering the range of hours selected
In the advance report interface users are able of:
1. Selecting one or many days of the weak.
2. Selecting a range of dates and / or hours.
3. Adding user comments.
4. Observing tables with the data retrieved and reports previously generated.
This software was developed for industrial service and academic use. If you like to perform a field test, please contact me:
Prof. Bladimir Bacca Cortes Ph.D.
Address: Cra. 100, Street 13, Universidad del Valle, Melendez, Building 354, Office 2006.
Tel: +5723212100 Ext. 7656