Method and system for providing parking service

A non-transitory computer-readable recording medium storing instructions that, when executed by at least one processor, cause the at least one processor to perform a method for providing a parking service, the method comprising:

• obtaining a parking monitoring image from a sensor monitoring a parking area of a parking lot;
• processing the parking monitoring image using the at least one processor to determine whether a target vehicle is parked in a specific parking area, the processing comprising:
• calculating an overlap degree of the target vehicle and the specific parking area by analyzing the parking monitoring image obtained from a sensor at a first time and determining whether the target vehicle is parked based on the calculated overlap degree;
• recording the target vehicle in an exit queue based on an analysis of a subsequent parking monitoring image indicating that the target vehicle has left the specific parking area;
• performing an exit process for the target vehicle from the exit queue based on detecting, from a further subsequent parking monitoring image, that the target vehicle is moving to exit the specific parking area;
• generating a feature vector by applying appearance of the detected vehicle to a machine learning algorithm or a deep learning algorithm, and using the generated feature vector to discriminate whether the detected vehicle and the target vehicle included in the exit queue are the same vehicle;
• and
• automatically performing a payment process for the exit of the target vehicle based on preregistered payment information associated with the target vehicle when the target vehicle is parked,
• wherein the feature vector includes at least one of a global feature based on appearance of the detected vehicle and a local feature.

A server comprising:

• a communication unit;
• a memory configured to store at least one command;
• a processor configured to execute the command stored in the memory,
• wherein the command executed by the processor includes:
• calculating an overlap degree of a target vehicle and a specific parking area by analyzing a parking monitoring image obtained from a sensor for monitoring the specific parking area of a parking lot at a first time and determining whether the target vehicle is parked based on the calculated overlap degree;
• recording the target vehicle in an exit queue when analysis of a parking monitoring image, obtained from a sensor at a second time, indicates that the target vehicle departs the specific parking area;
• performing an exit process when analysis of a parking monitoring image, obtained from a sensor at a third time, detects that the target vehicle in the exit queue is moving to leave the specific parking area;
• generating a feature vector by applying appearance of the detected vehicle to a machine learning algorithm or a deep learning algorithm, and using the generated feature vector to discriminate whether the detected vehicle and the target vehicle included in the exit queue are the same vehicle;
• and
• performing a payment process in accordance with an exit of the target vehicle based on payment information pre-registered when the target vehicle is parked,
• wherein the feature vector includes at least one of a global feature based on appearance of the detected vehicle and a local feature.