Cross-sensor object-space correspondence analysis method and system using same

A cross-sensor object-space correspondence analysis method for detecting at least one object in a space by using cooperation of a plurality of image sensing devices, the method being implemented by an edge computing architecture including a plurality of information processing units in the image sensing devices respectively and a main information processing device, and the method including:

• the information processing units transmitting detected data to the main information processing device, the detected data being raw data of plural frames of image sensed by the image sensing devices, or at least one local grid code or at least one global grid code generated by using a first inference process to process the raw data, and the main information processing device using the detected data to determine a projection point on a reference plane corresponding to the space for each of the at least one object, where each of the image sensing devices has an image plane, the raw data, each of the at least one local grid code, and each of the at least one global grid code all correspond to a time record, the first inference process includes:
• performing a target locating procedure on the raw data to locate at least one pixel position in a frame of the image corresponding to at least one of the at least one object at a time point;
• and using a first grid code look-up table to perform a first mapping operation on each of the at least one pixel position to generate the at least one local grid code corresponding to at least one of the grids of one of the image planes, or using a second grid code look-up table to perform a second mapping operation on each of the at least one pixel position to generate the at least one global grid code corresponding to at least one of the grids of the reference plane;
• and
• the main information processing device performing a second inference procedure on the raw data of the frames of image provided by the information processing units to generate at least one of the at least one global grid code, and using the global grid code to represent the projection point;
• or using a code-code look-up table to perform a third mapping operation on the at least one local grid code provided by each of the information processing units to obtain at least one of the global grid codes to represent the projection point of the at least one object on the reference plane;
• or using the global grid code provided by the information processing units to represent the projection point, where the second inference procedure includes:
• performing an object positioning procedure on the raw data to locate at least one pixel position in a frame of the image corresponding to at least one of the at least one object;
• and using a second grid code look-up table to perform the second mapping operation on each of the at least one pixel position to generate at least one of the global grid codes corresponding to at least one of the grids of the reference plane.