Image processing using photonic quantum computing

1 . An apparatus comprising:

a processor of a converter system configured to:

receive image data, wherein the image data comprises red-channel image data, green-channel image data, and blue-channel image data, and wherein the image data is represented by classical binary bits; and

convert the image data to a photon beam comprising converted image data, wherein the converted image data comprises converted red-channel image data, converted green-channel image data, and converted blue-channel image data, and wherein the converted image data is represented by photonic quantum bits;

a photonic quantum computing system communicatively coupled the converter system, the photonic quantum computing system comprising:

a beam splitter configured to:

receive the photon beam; and

split the photon beam into a red-channel photon beam, a green-channel photon beam, and a blue-channel photon beam, wherein the red-channel photon beam comprises the red-channel image data, the green-channel photon beam comprises the green-channel image data, and the blue-channel photon beam comprises to the blue-channel image data; and

a photonic quantum processor coupled to the beam splitter and implementing a quantum neural network, wherein the quantum neural network comprises a plurality of quantum neural network clusters, and wherein the photonic quantum processor is configured to:

receive the red-channel photon beam, the green-channel photon beam, and the blue-channel photon beam; and

for each of the red-channel photon beam, the green-channel photon beam, and the blue-channel photon beam, processing in parallel the red-channel photon beam, the green-channel photon beam, and the blue-channel photon beam by a respective quantum neural network cluster of the plurality of quantum neural network clusters by:

(a) converting a converted N-channel image data into grayscale image data, wherein the converted N-channel image data comprises the converted red-channel image data, the converted green-channel image data, or the converted blue-channel image data;

(b) applying one or more first filters to the grayscale image data to extract from the grayscale image data a plurality of target image features, wherein the extracted plurality of target image features comprises a plurality of text characters of one or more text sequences included in the grayscale image data;

(c) removing first orientation effects associated with the extracted plurality of target image features by inputting the extracted plurality of target image features to first pooling layers;

(d) subsequent to removing the first orientation effects, refining the extracted plurality of target image features by applying one or more second filters to the extracted plurality of target image features;

(e) in response to refining the extracted plurality of target image features, removing second orientation effects associated with the refined and extracted plurality of target image features by inputting the refined and extracted plurality of target image features to second pooling layers;

(f) comparing the refined and extracted plurality of target image features to a plurality of pre-trained features extracted from pre-training images;

(g) in response to the refined and extracted plurality of target image features matching respective pre-trained features, identifying a plurality of identified image features; and

(h) sending the plurality of identified image features to the converter system as an N-channel photon beam, wherein the N-channel photon beam comprises a second red-channel photon beam, a second green-channel photon beam, or a second blue-channel photon beam.

2 . The apparatus of claim 1 , wherein the processor of the converter system is further configured to:

receive the second red-channel photon beam, the second green-channel photon beam, and the second blue-channel photon beam; and

convert the plurality of identified image features to the plurality of converted identified image features, wherein the plurality of converted identified image features are represented by classical binary bits.

3 . The apparatus of claim 2 , further comprising:

a processor of a data extraction system configured to:

generate extracted data by combining the plurality of converted identified image features for each of the second red-channel photon beam, the second green-channel photon beam, and the second blue-channel photon beam.

4 . The apparatus of claim 1 , wherein each quantum neural network cluster of the plurality of quantum neural network clusters comprises a plurality of quantum neurons.

5 . A method comprising:

receiving image data, wherein the image data comprises red-channel image data, green-channel image data, and blue-channel image data, and wherein the image data is represented by classical binary bits;

converting the image data to a photon beam comprising converted image data, wherein the converted image data comprises converted red-channel image data, converted green-channel image data and converted blue-channel image data, and wherein the converted image data is represented by photonic quantum bits;

splitting the photon beam into a red-channel photon beam, a green-channel photon beam, and a blue-channel photon beam, wherein the red-channel photon beam comprises the red-channel image data, the green-channel photon beam comprises the green-channel image data, and the blue-channel photon beam comprises to the blue-channel image data;

receiving the red-channel photon beam, the green-channel photon beam, and the blue-channel photon beam by a quantum neural network, wherein the quantum neural network comprises a plurality of quantum neural network clusters; and

for each of the red-channel photon beam, the green-channel photon beam, and the blue-channel photon beam, processing in parallel the red-channel photon beam, the green-channel photon beam, and the blue-channel photon beam by a respective quantum neural network cluster of the plurality of quantum neural network clusters by:

(a) converting a converted N-channel image data into grayscale image data, wherein the converted N-channel image data comprises the converted red-channel image data, the converted green-channel image data, or the converted blue-channel image data;

(b) applying one or more first filters to the grayscale image data to extract from the grayscale image data a plurality of target image features, wherein the extracted plurality of target image features comprises a plurality of text characters of one or more text sequences included in the grayscale image data;

(c) removing first orientation effects associated with the extracted plurality of target image features by inputting the extracted plurality of target image features to first pooling layers;

(d) subsequent to removing the first orientation effects, refining the extracted plurality of target image features by applying one or more second filters to the extracted plurality of target image features;

(e) in response to refining the extracted plurality of target image features, removing second orientation effects associated with the refined and extracted plurality of target image features by inputting the refined and extracted plurality of target image features to second pooling layers;

(f) comparing the refined and extracted plurality of target image features to a plurality of pre-trained features extracted from pre-training images;

(g) in response to the refined and extracted plurality of target image features matching respective pre-trained features, identifying a plurality of identified image features; and

(h) sending the plurality of identified image features to the converter system as an N-channel photon beam, wherein the N-channel photon beam comprises a second red-channel photon beam, a second green-channel photon beam, or a second blue-channel photon beam.

6 . The method of claim 5 , further comprising:

receiving the second red-channel photon beam, the second green-channel photon beam, and the second blue-channel photon beam; and

converting the plurality of identified image features to the plurality of converted identified image features, wherein the plurality of converted identified image features are represented by classical binary bits.

7 . The method of claim 6 , further comprising:

generating extracted data by combining the plurality of converted identified image features for each of the second red-channel photon beam, the second green-channel photon beam, and the second blue-channel photon beam.

8 . The method of claim 5 , wherein each quantum neural network cluster of the plurality of quantum neural network clusters comprises a plurality of quantum neurons.

9 . A non-transitory computer-readable medium storing instructions that, when executed by one or more processors, cause the one or more processors to:

receive image data, wherein the image data comprises red-channel image data, green-channel image data, and blue-channel image data, and wherein the image data is represented by classical binary bits;

convert the image data to a photon beam comprising converted image data, wherein the converted image data comprises converted red-channel image data, converted green-channel image data and converted blue-channel image data, and wherein the converted image data is represented by photonic quantum bits;

split the photon beam into a red-channel photon beam, a green-channel photon beam, and a blue-channel photon beam, wherein the red-channel photon beam comprises the red-channel image data, the green-channel photon beam comprises the green-channel image data, and the blue-channel photon beam comprises to the blue-channel image data;

receive the red-channel photon beam, the green-channel photon beam, and the blue-channel photon beam by a quantum neural network, wherein the quantum neural network comprises a plurality of quantum neural network clusters; and

for each of the red-channel photon beam, the green-channel photon beam, and the blue-channel photon beam, process in parallel the red-channel photon beam, the green-channel photon beam, and the blue-channel photon beam by a respective quantum neural network cluster of the plurality of quantum neural network clusters by:

(a) converting a converted N-channel image data into grayscale image data, wherein the converted N-channel image data comprises the converted red-channel image data, the converted green-channel image data, or the converted blue-channel image data;

(b) applying one or more first filters to the grayscale image data to extract from the grayscale image data a plurality of target image features, wherein the extracted plurality of target image features comprises a plurality of text characters of one or more text sequences included in the grayscale image data;

(c) removing first orientation effects associated with the extracted plurality of target image features by inputting the extracted plurality of target image features to first pooling layers;

(d) subsequent to removing the first orientation effects, refining the extracted plurality of target image features by applying one or more second filters to the extracted plurality of target image features;

(e) in response to refining the extracted plurality of target image features, removing second orientation effects associated with the refined and extracted plurality of target image features by inputting the refined and extracted plurality of target image features to second pooling layers;

(f) comparing the refined and extracted plurality of target image features to a plurality of pre-trained features extracted from pre-training images;

(g) in response to the refined and extracted plurality of target image features matching respective pre-trained features, identifying a plurality of identified image features; and

(h) sending the plurality of identified image features to the converter system as an N-channel photon beam, wherein the N-channel photon beam comprises a second red-channel photon beam, a second green-channel photon beam, or a second blue-channel photon beam.

10 . The non-transitory computer-readable medium of claim 9 , wherein the instructions, when executed by the one or more processors, further cause the one or more processors to:

receive the second red-channel photon beam, the second green-channel photon beam, and the second blue-channel photon beam; and

convert the plurality of identified image features to the plurality of converted identified image features, wherein the plurality of converted identified image features are represented by classical binary bits.

11 . The non-transitory computer-readable medium of claim 9 , wherein the instructions, when executed by the one or more processors, further cause the one or more processors to:

generate extracted data by combining the plurality of converted identified image features for each of the second red-channel photon beam, the second green-channel photon beam, and the second blue-channel photon beam.

12 . The non-transitory computer-readable medium of claim 9 , wherein at least one of the one or more processors is a photonic quantum processor.