The Hong Kong University of Science and Technology

Leland Stanford Junior University

Shenzhen Nanobioimaging Limited

Shengwang Du

Teng Zhao

Wei Wang

Shuhuai Yao

Chinmay Belthangady

Ying Wang

Ying Zuo

Shanchao Zhang

You-Chiuan Chen

Shuyu Zhou

Yumian Su

Guang Yu Yin

Rong Chen

Yue Jiang

Yujun Zhao

Junwei Liu

Ming Tak Michael Loy

Sze Cheung Lau

Michael M T Loy

Bohan Li

The present application provides methods, devices, systems and non-transitory computer readable storage media for image processing. In an aspect, there is provided a computer-implemented method of image processing, the method comprising: receiving a low resolution image of an object; generating an edge map of the low resolution image by an edge extractor; and inputting the edge map and the low resolution image to a neural network to reconstruct a super resolution image of the object, wherein the neural network is trained using a multicomponent loss function.

Deep learning based single frame super resolution microscopy image processing

An all-optical neural network that utilizes light beams and optical components to implement layers of the neural network is disclosed herein. The all-optical neural network includes an input layer, zero or more hidden layers, and an output layer. Each layer of the neural network is configured to simulate linear and nonlinear operations of a conventional artificial neural network neuron on an optical signal. In an embodiment, the optical linear operation is performed by a spatial light modulator and an optical lens. The optical lens performs a Fourier transformation on the set of light beams and sums light beams with similar propagation orientations. The optical nonlinear operation is implemented utilizing a nonlinear optical medium having an electromagnetically induced transparency characteristic whose transmission of a probe beam of light is controlled by the intermediate output of a coupling beam of light from the optical linear operation.

All optical neural network

Methods and systems for generating non-diffracting light sheets for multicolor fluorescence microscopy are disclosed. A method for generating a non-diffracting light patterned Bessel sheet comprises transmitting an input light beam through a Fourier transform lens the input light beam has a spatial intensity pattern at a first plane, and a Fourier plane is formed after the Fourier transform lens to obtain a first light beam; transmitting the first light beam through an annulus mask to obtain a second light beam; and transmitting the second light beam through an excitation objective lens to form a non-diffracting patterned light sheet. A method for generating a non-diffracting light line Bessel sheet comprises transmitting an input light beam at a first lane through an annulus mask to obtain a first light beam; and transmitting the first light beam through an excitation objective lens to form a non-diffracting Bessel light sheet.

Methods and systems for generating non-diffracting light sheets for multicolor fluorescence microscopy

A two-color fluorescence localized super-resolution biological microscopy method and system are disclosed. The method includes: performing two-color fluorescence labeling on a biological sample by using Alexa647 and Alexa750 fluorescent molecules or Cy5 and Cy7 fluorescent molecules, and soaking the biological sample in an imaging buffer solution; illuminating the biological sample by laser to generate a flashing fluorescent signal of a first channel and a flashing fluorescent signal of a second channel respectively; constructing a super-resolution image of a first biological structure and a super-resolution image of a second biological structure respectively; and aligning the super-resolution image of the first biological structure and the super-resolution image of the second biological structure so as to construct a super-resolution image of a third biological structure. Thereby, the present method and system do not generate channel crosstalk and can significantly reduce background noise, so that the imaging quality is improved.

Two-color fluorescence localization super-resolution biological microscopy method and system

A two-dimensional (2D) magneto-optical trap (MOT) for alkali neutral atoms establishes a zero magnetic field along the longitudinal symmetry axis. Two of three pairs of trapping laser beams do not follow the symmetry axes of the quadruple magnetic field and are aligned with a large non-zero degree angles to the longitudinal axis. In a dark-line 2D MOT configuration, there are two orthogonal repumping beams. In each repumping beam, an opaque line is imaged to the longitudinal axis, and the overlap of these two line images creates a dark line volume in the longitudinal axis where there is no repumping light. The zero magnetic field along the longitudinal axis allows the cold atoms maintain a long ground-state coherence time without switching off the MOT magnetic field, which makes it possible to operate the MOT at a high repetition rate and a high duty cycle.

Two-dimensional magneto-optical trap for neutral atoms

Growing community of inventors

Hong Kong, China

Shengwang Du