Facial recognition system, facial recognition server, and facial recognition method

A facial recognition server which is connected to a camera device, the server comprising:

• a feature extractor, which in operation, extracts features of facial image data including a face of a person shown in video image data, based on the video image data from the camera device;
• a preservation unit, which in operation, preserves the features of the facial image data extracted by the feature extractor, in a facial feature memory;
• and
• a statistical processing unit, which in operation:
• performs statistical processing by using the features of the facial image data extracted by the feature extractor, in a case where:
• the features of facial image data extracted by the feature extractor have a similarity that is equal to or greater than a predetermined value to features of facial image data that are present in the facial feature memory, and a predetermined time has elapsed from an imaging time point by the camera device of the video image data that includes the facial image data,
• wherein the facial recognition server discards the features of the facial image data extracted by the feature extractor, in a case where:
• the features of facial image data extracted by the feature extractor have a similarity that is equal to or greater than a predetermined value to features of facial image data that are present in the facial feature memory, the predetermined time has not elapsed from an imaging time point by the camera device of the video image data that includes the facial image data, and without satisfying one of preservation conditions:
• the facial image data is from the first captured facial image of the person, and an orientation of a face included in the facial image data is largely shown front face.

A facial recognition method in a facial recognition system in which a camera device and a facial recognition server are connected to each other, the method comprising executing at least in part using a processor, at least the following processing:

• processing of extracting features of facial image data including the face of a person shown in video image data, based on the video image data obtained by the camera device imaging an imaging area;
• preserving the extracted features of the facial image data in a facial feature memory;
• performing statistical processing by using the extracted features of the facial image data in a case where:
• the extracted features of facial image data have a similarity that is equal to or greater than a predetermined value to features of stored facial image data that are present in the facial feature memory, and a predetermined time has elapsed from an imaging time point of the video image data that includes the facial image data;
• and
• discarding the extracted features of the facial image data in a case where:
• the extracted features of facial image data have a similarity that is equal to or greater than a predetermined value to stored features of facial image data that are present in the facial feature memory, the predetermined time has not elapsed from an imaging time point of the video image data that includes the facial image data, and without satisfying one of preservation conditions:
• the facial image data is from the first captured facial image of the person, and an orientation of a face included in the facial image data is largely shown front face.