It’s almost hopeless to find a parking place on streets of Kyiv, Ukrainian capital city. Yet it’s possible to incure a millions of losses from a municipal parking buiseness for years in row. We decided to figure out what is real situation on this market, and compared official financial statements about parking lots with real picture(s).
Most of the drivers do not use parking machines. Instead they could hand a twenty hryvnias (about 1 USD) for a parking assistant, without asking for a receipt. Next, parking worker at the end of a day could pay with a small part of cash through the parking machine, but take the rest of money for himself.
Thus, we decided to assess how much money one might actually collect from a downtown parking lots. So we recorded some of the parking places on video and then calculated the number of cars (and possible income from parking fee).
Next we compared our results with open-data financial statements from the KPTS (community-owned enterprise in charge of parking places) for these parking lots for the same period of time.
The difference is huge: for all three parking lots we recorded on video the income could be 5-10 times bigger then officially reported amounts
What makes this project innovative?
We also used another way to present visual evidence. It was made by selecting the single line of pixels which crossed parking zone on each video frame. Next we "stitched" such lines together in one image, for the working time of each parking. We called such a photo visualization "compressed parking time" --- when couple of hours of video represented as one image. You could see one of examples as a featured image in this application entry.
Journalism: For the first time we obtained an estimation how much money city are loosing on municipal parking places in form of missed income.
What was the impact of your project? How did you measure it?
Source and methodology
Official financial results for these parking places were obtained from open-data portal of city council (http://gis.kievcity.gov.ua/)
Finally, we compared these two numbers for each day on each parking lot, and presented their along with a video evidence about number of parked cars.
To read more about our methodology please see this detailed article on Medium, in English: https://email@example.com/parkings-title-c988ba6241f4
OpenCV and Python to get data from video and create "compressed parking time" photo viz
Vlad Gerasymenko - web development and JS programming
Yevheniia Drozdova - web development
Nadja Kelm - design
Anatoliy Bondarenko - OpenCV programming
Maria Klokova - data analysis