The topic of our last posts was “Urban Analytics module” and the tools it includes. One of the tools is “Road accidents” for revealing the road accidents hotspots.
Now let us introduce “Visualizing Vehicle Distribution in a City” tool. It is characterized by its main purpose: visualization of car density and their location in the city. Knowing the detailed information of this kind can be useful in many processes like planning the layout of parking spots or evaluating car-dependency level of the city.
When a polygonal object representing real-life objects like green spaces or lawns is stored in the system, it is possible to reveal areas where lawn parking violations take place.
But such information is not easy to obtain. The proposed method for solving the problem is based on computer vision technology. The neural network takes orthophotomaps or high-resolution satellite images and distinguishes vehicles on it, classifying each as a car or a truck. A centroid is then assigned to each vehicle as its representation in the system.
As a result, thematic data layer is created that can be interpreted by GIS tools. Use cases include comparing with other objects of the system to solve real-life issues.
The accumulation of historical data in the system with new relevant orthophotomaps becoming available will help to see the dynamics of the situation.
The tool also allows you to calculate the density of cars in each point of the city and display the results of the calculation on the map. Following map shows distribution of cars and cars clusters around the city.
Another developed feature is identification of residential areas with overcrowded parking lots.
Announcing the topic of next time: the “Number of residents” tool.
You can read about all the features of the Urban Analytics module in our article.