US 12,387,364 B2
Methods and apparatus of brain-like in-pixel intelligent processing system
Tuan A Duong, La Verne, CA (US); and Quangdao Duong, La Verne, CA (US)
Assigned to ADAPTIVE COMPUTATION LLC., La Verne, CA (US)
Filed by ADAPTIVE COMPUTATION LLC, La Verne, CA (US)
Filed on Jun. 5, 2023, as Appl. No. 18/205,643.
Prior Publication US 2024/0404096 A1, Dec. 5, 2024
Int. Cl. G06T 7/70 (2017.01); G06V 10/25 (2022.01); G06V 10/44 (2022.01); G06V 10/82 (2022.01)
CPC G06T 7/70 (2017.01) [G06V 10/25 (2022.01); G06V 10/44 (2022.01); G06V 10/82 (2022.01); G06V 2201/07 (2022.01)] 20 Claims
OG exemplary drawing
 
1. An in-pixel processing array for a brain-like in-pixel intelligent processing system, comprising:
a plurality of in-pixel processing units, wherein the in-pixel processing array forms an image acquisition and bio-inspired processing imager configured to processes raw gray information in a pixel level; wherein each of the in-pixel processing units comprises:
a photogate sensor, wherein the photogate sensor captures a pixel of the image of an object and produces an Iout current to the in-pixel processing array;
an average circuit, wherein the average circuit receives and averages Iout current from the in-pixel processing unit and a plurality of neighboring in-pixel processing units through a plurality of input channels, and the averaged Iout current from the plurality of neighboring in-pixel processing units is used as saccadic eye movements to generate a periphery (P) element;
a subtraction circuit, wherein the periphery (P) element generated is used by the subtraction circuit to generate a fovea (F) element of the in-pixel processing unit; and
an absolute circuit, wherein F element generated in pixel level is sent to the absolute circuit via pixel mapping to generate lateral geniculate nucleus (LGN) element of the in-pixel processing unit,
wherein the P element, the F element, and the LGN element from each of in-pixel processing unit of the in-pixel processing array are combined at a feature vector generator to form a P feature vector, an F feature vector, and an LGN feature vector of the in-pixel processing array, the P feature vector, the F feature vector, and the LGN feature vector are used by a neural network processor to detect the object, to identify the object, and to determine the location of the object, and a region and object of interest identifier identifies a region and object of interest from the object detected and identified by the neural network processor and provide feedback of the identified region and object of interest and the processed P, F, and LGN feature vectors of the brain-like in-pixel intelligent processing system to the in-pixel processing array to improve the object detection and identification.