Cross-sensor object-attribute analysis method and system

A cross-sensor object-attribute analysis method, which is applied in a space disposed with a plurality of image sensing devices to enable the image sensing devices to cooperatively detect at least one object, and is realized by an edge computing architecture, the edge computing architecture including a main information processing device and a plurality of information processing units respectively disposed in the image sensing devices, the method including:

• the information processing units transmitting detected data to the main information processing device, the detected data being raw data of sensed multiple images of the image sensing devices or data of one or more attribute vectors generated by processing the raw data using an inference process, and the main information processing device using the one or more attribute vectors to correspond to the at least one object, where each of the image sensing devices has an image plane, and the raw data and each of the attribute vectors all correspond to a time record, the inference process including:
• performing an object recognition procedure on the raw data to define a bounding box for at least one of the at least one object;
• and using a first AI module to perform an attribute evaluation calculation on each of one or more of the bounding boxes to generate one or more of the attribute vectors corresponding to one or more of the objects appearing in one of the images at a time point;
• and
• the main information processing device performing the inference process on the raw data of the images provided by the information processing units to generate one or more of the attribute vectors, and using each of the attribute vectors to correspond to one of the at least one object, or the main information processing device directly using the one or more of the attribute vectors provided by the information processing units to correspond to the at least one subject.