Lensless near-contact imaging system for micro assembly

Patent metadata

Field	Value
Publication number	US-12477237-B2
Application number	US-202318529416-A
Country	US
Kind code	B2
Filing date	Dec 5, 2023
Priority date	Dec 5, 2023
Publication date	Nov 18, 2025
Grant date	Nov 18, 2025

Abstract

A machine vision system and method use lensless near-contact imaging with coherent illumination, or incoherent illumination, and high pixel count large format sensors (e.g., equivalent to at least 20 to 65 mega-pixels) to produce diffraction patterns of the micro-objects or the gray scale images of the micro-objects over a large overall field-of-view of the machine vision system. The machine vision system provides feedback to a microassembler system to position, orient, and assemble microscale devices like micro-LEDs over large working areas. The effective resolution of the machine vision system can be further improved through the use of gray scale and super-resolution image processing techniques.

First claim

What is claimed is:

1 . A method for operating a machine vision system suitable for use with a microassembler system for inspection of assembly of micro-objects on a planar working surface, the method comprising:
arranging an array of lensless near-contact image-capture modules (LNCIM) vertically proximate to a working optical inspection region on a planar working surface of a transparent substrate, wherein each LNCIM having a high pixel-count large format image sensor (image sensor) vertically proximate and near-contact to and facing the planar working surface and being associated with a LNCIM field-of-view (FOV) region on the planar working surface; arranging one or more illumination light sources, optically coupled to respective one or more source optical trains, located and directed to pass and emit illumination light, in a defined wavelength range in a near infrared wavelength range, from the one or more source optical trains toward the planar working surface of the transparent substrate, the one or more illumination light sources optically coupled to respective one or more source optical trains being located on one side of the transparent substrate and the array of lensless near-contact image-capture modules being located on an opposite side of the transparent substrate; selectively turning ON the one or more illumination light sources optically coupled to the respective one or more source optical trains thereby preferentially passing and directing emitted electromagnetic radiation (light) in the defined wavelength range in the near infrared wavelength range to illuminate the working optical inspection region on the planar working surface; receiving, by the image sensor of each LNCIM in the array while the illumination light source is ON, light signals from each LNCIM FOV region associated with each respective image sensor of an LNCIM in the array, wherein received light signals from a micro-object disposed in a LNCIM FOV region on the planar working surface comprise a diffraction pattern received by the associated image sensor of an LNCIM in the array; capturing, by the image sensor of each LNCIM in the array, a LNCIM-captured image of the light signals received from the associated LNCIM FOV region on the planar working surface, wherein the LNCIM-captured image being associated with a LNCIM-captured image resolution; adjusting the LNCIM-captured image of at least one LNCIM in the array; detecting at least one diffraction pattern in the LNCIM-captured image associated with an LNCIM FOV region on the planar working surface, the at least one diffraction pattern corresponding to at least one micro-object disposed in the LNCIM FOV region; utilizing image processing of the diffraction pattern to determine at least one of a type, location, centroid, horizontal orientation, or vertical orientation, of the at least one micro-object; comparing the detected diffraction pattern in the LNCIM-captured image to models of diffraction patterns associated with features of known micro-objects; identifying, based on the comparing, at least one of a location, a horizontal orientation, a vertical orientation, a centroid, or a type, of a micro-object in the LNCIM FOV region associated with the LNCIM-captured image; and generating, based on the identifying, captured-image data associated with the identified at least one of a location, horizontal orientation, vertical orientation, centroid, or type, of the at least one micro-object disposed in the LNCIM FOV region, for providing the captured-image data to a microassembler system.

2 . The method of claim 1 , wherein the comparing includes comparing irradiance levels of light signals in the at least one diffraction pattern in the LNCIM-captured image to irradiance levels of light signals in the models of diffraction patterns.

3 . The method of claim 1 , wherein the one or more illumination light sources are coherent illumination light sources that emit coherent illumination light, and wherein a plurality of LNCIM FOV regions on the planar working surface, respectively associated with the array of LNCIM, form a working FOV region on the planar working surface for the machine vision system, the method comprising:
adjusting the LNCIM-captured image of at least one LNCIM in the array, wherein the adjusting includes adjusting a LNCIM-captured image resolution; and stitching together a plurality of LNCIM-captured images of adjacent LNCIM in the array; and forming, based on the plurality of LNCIM-captured images stitched together, a working FOV image of the working FOV region associated with a working FOV image resolution for the machine vision system, wherein the working FOV region is larger than each LNCIM FOV region in the plurality and the working FOV image resolution is at least equal to or greater than each LNCIM-captured image resolution respectively associated with each LNCIM FOV region in the plurality.

4 . The method of claim 3 , wherein the stitching comprises side-by-side stitching of LNCIM-captured images of adjacent LNCIM FOV regions to form the working FOV image of the working FOV region.

5 . The method of claim 3 , wherein the stitching comprises feathered stitching of LNCIM-captured images of adjacent LNCIM FOV regions to form the working FOV image of the working FOV region.

6 . The method of claim 3 , wherein the stitching comprises staggered stitching of LNCIM-captured images of adjacent LNCIM FOV regions to form the working FOV image of the working FOV region.

7 . The method of claim 6 , wherein the staggered stitching comprises a stitching geometry based on a step-and-repeat assembly process to stitch LNCIM-captured images of adjacent LNCIM FOV regions.

8 . The method of claim 7 , wherein the step-and-repeat assembly process comprises:
a working FOV region arranged in rows and columns of LNCIM FOV regions, where the LNCIM in the array capture images of their respective LNCIM FOV regions according to:
a horizontal stagger pitch (P H ) of LNCIM FOV regions in a same row,
a vertical stagger pitch (P V ) of LNCIM FOV regions in separate rows, and
n is a total number of rows, and
s is a total number of steps, and
wherein the step-and-repeat assembly process achieves an efficiency of capturing LNCIM-captured images, and stitching adjacent LNCIM-captured images based on a formula:
=number of rows divided by number of steps
= n /( P H ( P V +n )).

9 . The method of claim 8 , wherein P H equals P V equals 1, n is a number greater than or equal to 100, and the efficiency is approximately 1, within a tolerance of plus or minus one tenth.

10 . The method of claim 3 , wherein the stitching comprises staggered stitching of LNCIM-captured images of adjacent LNCIM FOV regions, based on a step-and-repeat assembly process to stitch LNCIM-captured images of adjacent LNCIM FOV regions, thereby forming the working FOV image of the working FOV region; and
wherein the step-and-repeat assembly process comprises one of:
holding the transparent substrate stationary while moving the LNCIM array to a new position and an image sensor of at least one LNCIM of the LNCIM Array capturing from an associated LNCIM FOV region an LNCIM-captured image at the new position of the LNCIM Array; or
holding the LNCIM array stationary while moving the transparent substrate to a new position and an image sensor of at least one LNCIM of the LNCIM Array capturing from an associated LNCIM FOV region an LNCIM-captured image at the new position of the transparent substrate.

11 . The method of claim 10 , wherein the step-and-repeat assembly process comprises:
holding the transparent substrate stationary while moving, in a synchronized movement w

Assignees

• Palo Alto Res Ct Inc
• Xerox Corp

Inventors

• Maeda Patrick Yasuo
• Ping Lu Jeng
• Chow Eugene

Classifications

• G06T7/97

  Determining parameters from multiple pictures (depth or shape recovery from multiple images G06T7/55; stereo camera calibration G06T7/85) · CPC title

• G06T2207/20221

  Image fusion; Image merging · CPC title

• G06V2201/07

  Target detection · CPC title

• B81C99/002

  Apparatus for assembling MEMS, e.g. micromanipulators (micromanipulators per se B25J7/00) · CPC title

• G06V10/44 Primary

  Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components · CPC title