# US 12,007,277 B2 · Hyperspectral imager method and apparatus

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 12,007,277 B2
- Document ID: US12007277B2
- Status: Granted
- Kind: B2
- Application number: 17/412,029
- Filed: 2021-08-25
- Published: 2022-02-03
- Granted: 2024-06-11
- Earliest claimed: 2015-01-14
- Assignees: Technological Resources Pty. Limited
- Applicants: Technological Resources Pty. Limited
- Inventors: Richard J. Murphy; Arman Melkumyan; Anna Chlingaryan; Dai Bang Nguyen; Stuart Wishart; Alex Lowe; Steven Scheding
- Web page: https://patentgrep.com/patent/US12007277B2

## CPC

- G01J 3/2823: {Imaging spectrometer}
- G01J 3/0202: {Mechanical elements; Supports for optical elements}
- G01J 3/0208: {using focussing or collimating elements, e.g. lenses or mirrors; performing aberration correction}
- G01J 3/0237: {Adjustable, e.g. focussing}
- G01J 3/0264: {Electrical interface; User interface}
- G01J 3/0286: {Constructional arrangements for compensating for fluctuations caused by temperature, humidity or pressure, or using cooling or temperature stabilization of parts of the device; Controlling the atmosphere inside a spectrometer, e.g. vacuum}
- G01J 3/0291: {Housings; Spectrometer accessories; Spatial arrangement of elements, e.g. folded path arrangements}
- G01J 3/28: Investigating the spectrum (using colour filters G01J3/51)
- G06F 18/214: Generating training patterns; Bootstrap methods, e.g. bagging or boosting
- G06F 18/24: Classification techniques
- H04N 23/11: for generating image signals from visible and infrared light wavelengths
- H04N 23/51: Housings
- H04N 23/71: Circuitry for evaluating the brightness variation
- H04N 23/73: by influencing the exposure time
- G01J 2003/2826: {Multispectral imaging, e.g. filter imaging}
- G01J: MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY (light sources F21, H01J, H01K, H05B; investigating properties of materials by optical means G01N)
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION

## Abstract

A compact hyperspectral imager adapted to operate in harsh environments and to conduct post acquisition signal processing to provide automated and improved hyperspectral processing results is disclosed. The processing includes luminance and brightness processing of captured hyperspectral images, hyperspectral image classification and inverse rendering to produce luminance invariance image processing.

## Claims

### Claim 1

A method of processing a series of hyperspectral images in order to classify its constituent parts, the method including the steps of:
(a) deriving a non-stationary observation angle dependent probabilistic model having a series of parameters for the series of hyperspectral images;
(b) training the series of probabilistic model parameters on mineral samples obtained from artificial light reflectance measurements; and
(c) utilising the probabilistic model on hyperspectral imagery acquired from sampling geographical conditions under natural lighting conditions, to classify constituent parts of the hyperspectral imagery.

### Claim 2 (depends on 1)

A method as claimed in claim 1 wherein said probabilistic model comprises a non-stationary covariance function.

### Claim 3 (depends on 2)

A method as claimed in claim 2 wherein said probabilistic model comprises a non-stationary observation angle dependent covariance function (OADCF).

### Claim 4 (depends on 1)

A method as claimed in claim 1 wherein said probabilistic model includes a multi task Gaussian process.

### Claim 5 (depends on 1)

A method as claimed in claim 1 wherein said training step includes training the images on reflectance spectra obtained utilising artificial lighting.

### Claim 6 (depends on 1)

A method as claimed in claim 1 wherein said probabilistic model includes a multi task Gaussian process utilising a non stationary covariance function that is lumination invariant.

### Claim 7 (depends on 1)

A method as claimed in claim 1 wherein said probabilistic model is a multi task covariance function.

### Claim 8 (depends on 1)

A method as claimed in claim 1 wherein said probabilistic model is derived from a portion of said hyperspectral imagery that include low levels of atmospheric absorption.