Automatic training of a parked vehicle detector for large deployment

An image processing system for training a parked vehicle detector, said system comprising:

• at least one image-capturing unit;
• a processor that communicates electronically with at least one image capturing unit;
• and
• a computer-usable medium embodying computer program code, said computer-usable medium capable of communicating with said processor, said computer program code comprising instructions executable by said processor and configured for:
• capturing video data of at least one parking site via at least one image-capturing unit associated with a parked vehicle detector training system for training said parked vehicle detector, said video data including an image of at least one vehicle and at least one parking space within an effective field of view;
• collecting a plurality of positive training samples from said video data captured by said at least one image-capturing unit based on combination of data including data provided by an automated image analysis and human-input auxiliary information provided by a crowd-sourcing unit that stores image and information with respect to a final position of at least one vehicle in a collected positive training sample collector;
• collecting a plurality of negative training samples from said video data based on said automated image analyses with respect to said captured video data;
• and
• utilizing said plurality of positive training samples and said plurality of negative training samples to train, re-train, and update said parked vehicle detector with respect to said at least one parking site for use in managing parking at said at least one parking site.

A non-transitory processor-readable medium storing code representing instructions to cause an image-processing process for training a parked vehicle detector, said code comprising instructions to

• capture video data of at least one parking site via at least one image-capturing unit associated with a parked vehicle detector training system for training said parked vehicle detector, said video data including an image of at least one vehicle and at least one parking space within an effective field of view;
• collect a plurality of positive training samples from said video data captured by said at least one image-capturing unit based on a combination of data including data provided by an automated image analysis and human-input auxiliary information provided by a crowd-sourcing unit that stores image and information with respect to a final position of at least one vehicle in a collected positive training sample collector;
• collect a plurality of negative training samples from said video data based on said automated image analyses with respect to said captured video data;
• and
• utilize said plurality of positive training samples and said plurality of negative training samples to train, re-train, and update said parked vehicle detector with respect to said at least one parking site for use in managing parking at said at least one parking site.