Neuromorphic bionic surface compound eye system for large field of view 3D motion detection

Abstract

The neuromorphic bionic surface compound eye system for large field of view 3D motion detection involves the field of bionic optics and 3D detection technology, and solves the problem that existing neuromorphic vision sensors are still planar vision sensors based on traditional CCD / CMOS architecture, which cannot be combined with bionic curved surfaces Issues such as the effective integration of compound eye optical systems, including bionic curved compound eye arrays, support substrates, neuromorphic curved surface detectors, data readout circuits, and data processing circuits; the supporting substrate is used to fix bionic curved surface compound eye arrays and neuromorphic curved surface detectors, The curved surface detector is composed of multiple neuromorphic detector elements that are sensitive to changes in light intensity caused by motion; the bionic curved surface compound eye array is composed of multiple lenses arranged on the curved surface; the lens focuses the light and passes through the support substrate. The light holes converge to the neuromorphic detection unit; the invention realizes real-time detection of three-dimensional dynamics in a large field of view, and has the advantages of all-weather, low power consumption, and high dynamic range.

Description

technical field [0001] The invention relates to the technical fields of bionic optics and three-dimensional detection, in particular to a neuromorphic bionic curved surface compound eye system for large field of view three-dimensional motion detection. Background technique [0002] Large field of view 3D dynamic real-time detection has an urgent need in the fields of automatic driving, robot / drone navigation, target recognition and tracking, surveillance and early warning. However, the traditional "frame"-based planar CCD / CMOS vision sensor has a small field of view, slow speed, and can only obtain two-dimensional information and lose depth information. In addition, the subsequent image processing is complicated, resulting in a large demand for resources and high power consumption, which cannot meet the needs of large-scale vision. Field three-dimensional dynamic real-time detection requirements. In order to solve this problem, new vision sensor research needs to be introdu...

AI Technical Summary

Problems solved by technology

However, this neuromorphic vision sensor is still a planar vision sensor based on the traditional CCD / CMOS architecture, which cannot be effectively integrated with the bionic surface compound eye optical system.

Images