US 10,796,190 B2Granted

Methods and apparatus for imaging of layers

AssigneesMassachusetts Institute of Technology, Georgia Tech Research Corporation

InventorsBarmak Heshmat Dehkordi, Albert Redo-Sanchez, Ramesh Raskar, Alireza Aghasi, Justin Romberg

Life of this patent

Estimating the term…

Loading patent…

Abstract

A sensor may measure light reflecting from a multi-layered object at different times. A digital time-domain signal may encode the measurements. Peaks in the signal may be identified. Each identified peak may correspond to a layer in the object. For each identified peak, a short time window may be selected, such that the time window includes a time at which the identified peak occurs. A discrete Fourier transform of that window of the signal may be computed. A frequency frame may be computed for each frequency in a set of frequencies in the transform. Kurtosis for each frequency frame may be computed. A set of high kurtosis frequency frames may be averaged, on a pixel-by-pixel basis, to produce a frequency image. Text characters that are printed on a layer of the object may be recognized in the frequency image, even though the layer is occluded.

Claims20

Dependent claims

A method comprising:

(a) illuminating with terahertz light an object that includes at least a first layer, a second layer and a third layer, the first, second and third layers being opaque in the visible spectrum and being occluded, in the visible spectrum, from direct view of a sensor;
(b) taking, at different times, measurements of terahertz light that has reflected from the first, second and third layers and is incident on the sensor, each measurement being a measurement of strength of an electric field at a pixel of the sensor at a particular time, the electric field being a function of intensity of incident terahertz light;
(c) identifying peaks in amplitude of a digital time-domain signal, which signal encodes the measurements;
and
(d) for each particular identified peak
(i) selecting a time window that includes a time at which the particular identified peak occurs,
(ii) calculating a discrete Fourier transform (DFT) of the time-domain signal in the time window,
(iii) calculating a set of frequency frames, in such a way that each particular frequency frame in the set is calculated for a particular frequency in the amplitude spectrum of the DFT, which particular frequency is different than that of any other frequency frame in the set,
(iv) calculating kurtosis of each frequency frame in the set,
(v) selecting a subset of the frequency frames, in such a way that kurtosis of each frequency frame in the subset exceeds a specified threshold, and
(vi) averaging the subset of frequency frames, on a pixel-by-pixel basis, to produce a frequency image for that particular identified peak;
wherein the method produces at least a first frequency image of the first layer, a second frequency image of the second layer, and a third frequency image of the third layer.

9 more elements

A method comprising:

(a) illuminating with infrared or ultraviolet light an object that includes at least a first layer, a second layer and a third layer, the first, second and third layers being opaque in the visible spectrum and being occluded, in the visible spectrum, from direct view of a sensor;
(b) taking, at different times, measurements of infrared or ultraviolet light that has reflected from the first, second and third layers and is incident on the sensor, each measurement being a measurement of incident light intensity at a pixel of the sensor at a particular time;
(c) identifying peaks in amplitude of a digital time-domain signal, which signal encodes the measurements;
and
(d) for each particular identified peak
(i) selecting a time window that includes a time at which the particular identified peak occurs,
(ii) calculating a discrete Fourier transform (DFT) of the time-domain signal in the time window,
(iii) calculating a set of frequency frames, in such a way that each particular frequency frame in the set is calculated for a particular frequency in the amplitude spectrum of the DFT, which particular frequency is different than that of any other frequency frame in the set,
(iv) calculating kurtosis of each frequency frame in the set,
(v) selecting a subset of the frequency frames, in such a way that kurtosis of each frequency frame in the subset exceeds a specified threshold, and
(vi) averaging the subset of frequency frames, on a pixel-by-pixel basis, to produce a frequency image for that particular identified peak,
wherein the method produces at least a frequency image of the first layer, a frequency image of the second layer, and a frequency image of the third layer.

9 more elements

An apparatus comprising:

(a) a light source configured to emit terahertz light;
(b) a sensor that is configured to take, at different times, measurements of light in the optical spectrum, each measurement being a measurement of incident light intensity at a pixel of the sensor at a particular time;
and
(c) one or more computers that are programmed
(i) to identify peaks in amplitude of a digital time-domain signal, which signal encodes the measurements, and
(ii) for each particular identified peak
(A) to select a time window that includes a time at which the particular identified peak occurs,
(B) to calculate a discrete Fourier transform (DFT) of the time-domain signal in the time window,
(C) to calculate a set of frequency frames, in such a way that each particular frequency frame in the set is calculated for a particular frequency in the amplitude spectrum of the DFT, which particular frequency is different than that of any other frequency frame in the set,
(D to calculate kurtosis of each frequency frame in the set,
(E) to select a subset of the frequency frames, in such a way that kurtosis of each frequency frame in the subset exceeds a specified threshold, and
(F) to average the subset of frequency frames, on a pixel-by-pixel basis, to produce a frequency image for that particular identified peak.

9 more elements

Dependencies are inferred from the claim text.

Facts

Application: 16/059,144Filed 9 Aug 2018 · us provisional application
Priority: 10 Aug 2017
Published: US 2019/0050669 A114 Feb 2019
Granted: 6 Oct 2020
Term ends: Not estimatedThe source records don’t support an estimate.

Classification6 codes

G06V 30/1429
Image acquisition: identifying or ignoring parts by sensing at different wavelengths
G06V 30/187
By deriving mathematical or geometrical properties from the whole image: frequency domain transformation; Autocorrelation
G06F 2218/14
Classification; Matching: by matching peak patterns
G06V 30/10
Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition: character recognition