Method and processing unit for correlating image data content from disparate sources
A signal processing appliance is disclosed that will simultaneously process the image data sets from disparate types of imaging sensors and data sets taken by them under varying conditions of viewing geometry, environmental conditions, lighting conditions, and at different times. Processing techniques that emulate how the human visual path processes and exploits data are implemented. The salient spatial, temporal, and color features of observed objects are calculated and cross-correlated over the disparate sensors and data sets to enable improved object association, classification and recognition. The appliance uses unique signal processing devices and architectures to enable near real-time processing.
1. A signal processing appliance configured to:
receive a plurality of image data sets observed at different times by a plurality of different sensor types each having a different viewing geometry;
assign a time stamp and hash identifier to each of the image data sets and synchronizing a parallel image data set input into a temporal processing, motion tracking and persistence tracking processing function and synchronizing the parallel image data set input into a separate convolution and object classification processing function;
simultaneously run at least one of a spatial, temporal, motion tracking, persistence tracking or color content convolution on the image data sets to identify one or more features of an object in the respective image data sets;
generate a plurality of sets of image data set metadata associated with the object feature identified in the respective image data sets;
perform a cross-modal correlation process on the respective sets of image data set metadata to correlate and identify the object feature in the image data sets;
and;
display the cross-correlated object on an electronic display.
2. The appliance of claim 1 configured to perform a correlation of a plurality of image scene objects based on an output of a plurality of Gabor filters of Higher Order Gaussian (HOG) processing.
3. The appliance of claim 1 configured to perform a process wherein a correlation of a plurality of image scene objects is performed based on an output of a plurality of Reichardt motion filters.
4. The appliance of claim 1 configured to perform a process wherein a correlation of image scene objects is performed based on an output of a multi-color filter bank.
5. The appliance of claim 1 configured to perform a Bayesian analysis to determine the object classifications.
6. The appliance of claim 1 configured to perform a plurality of simultaneous calculations of object features across disparate scenes and image data sets using multiple GPUs.