The performance is not adequate, what else can I configure?
WARNING: Please read the disclaimer to make sure that all of the installation constraints have been fulfilled!
To ensure proper functioning of AI-Parking the camera should be positioned so that the hidden sections of the vehicles to detect (it is not necessary to frame the entire vehicle) have a dimension of 30x30 pixels. The cameras should be installed so that the view of each parking spot is free of any occlusions caused by near vehicles.
In the drop-down menu on the left side, by selecting "Algorithm Parameters" and then “Traffic Monitoring” you can draw an area of interest directly on the live image on the left. The area of interest is a polygon with four vertices. You have to draw a sensor for each spot to be monitored. WARNING: to properly detect the occupancy (or lack of) of a parking spot, you should draw each area by not taking into consideration regions that might be occluded by near vehicles. Here some informations on how to set configuration parameters of this sensor:
Besides these two parameters it is also possible to configure the following to achieve an even better configuration: