HIGH THROUGHPUT INTERACTION MAPPING OF HUMAN PROTEOMES

Not Available SY Technology, Inc. proposes to develop a smart camera to automatically identify regions of interest (ROI) in imagery. The camera has two target recognition applications. First, to process a correlation surface (obtained by correlating an image with a filter) to identify, enhance and output ROIs for storage and further processing. The same camera can also delete obvious regions of no interest before imagery is input into the target recognition system. The desired on-board processing speed is achieved by utilizing a CMOS Pulse Coupled Neural Network (PCNN). The PCNN is selected for this high-speed camera for three reasons. First, this team has demonstrated that the PCNN segments imagery significantly better than thresholding methods. Secondly, a PCNN based target detection algorithm that identifies and deletes regions of no interest has been developed in prior work. Finally, SY Technology is fabricating a high speed CMOS PCNN chips for medical applications. Phase I will optimize a PCNN based algorithm to identify ROT in imagery. A camera of 512 x 512 pixels with 8-bit readout will be simulated/validated. Simulations will focus on speed, reliability and size. m e final design of the smart camera, processing > 60 fps, will be delivered. Option I proposes to fabricate a prototype. BENEFITS: In addition to obvious military applications, smart high speed cameras identifying regions of interest have many commercial applications. Medical image processing (radiology, patholog