Method and system of comparing digital labels

1 . A method of comparing at least two digital labels, the method comprising:

detecting, by a processor, a set of regions of interest (ROIs) corresponding to a set of objects in each of the at least two digital labels using a first Machine Learning (ML)/Deep Learning (DL) model;

classifying, by the processor, each of the set of objects into one of a set of text objects or a set of non-text objects using a second ML/DL model;

for each text object of the set of text objects,

extracting, by the processor, one or more key-value pairs of one or more attributes from the text object using a first Natural Language Processing (NLP) model; and

comparing, by the processor, the one or more key-value pairs and text data corresponding to the text object of the at least two digital labels using a second NLP model;

for each non-text object of the set of non-text objects,

categorizing, by the processor, the non-text object into an object category from a set of object categories using a third ML/DL model; and

comparing, by the processor, an ROI of the non-text object based on positional information and the object category in the at least two digital labels; and

generating, by the processor, an output report based on the comparison of the text data and the comparison of the ROI of each of the set of similar type of objects in the at least two digital labels.

2 . The method of claim 1 , further comprising detecting, by the processor, a boundary of each of the at least two digital labels using an image processing technique.

3 . The method of claim 1 , wherein the set of objects comprises symbols, logos, barcodes, QR-codes, data matrix, and text data, and wherein the set of object categories comprises symbols, logos, barcodes, data matrix, and QR-codes.

4 . The method of claim 3 , further comprising determining, by the processor, decoded data from an object in each of the at least two digital labels when the object category is one of the Barcodes, the Data Matrix, or the QR-codes.

5 . The method of claim 1 , further comprising:

for each of the at least two digital labels,

identifying, by the processor, address-associated non-text objects from the set of non-text objects;

extracting, by the processor, address-associated text objects from the set of text objects using a Named Entity Recognition (NER) model, wherein the address-associated text objects are within a predefined threshold proximity to the address-associated non-text objects;

creating, by the processor, masks corresponding to the address-associated text objects and address-associated non-text objects;

clustering, by the processor, address-associated ROIs based on the masks using image processing techniques; and

extracting, by the processor, address information from the address-associated ROIs using a third NLP model.

6 . The method of claim 1 , wherein extracting one or more key-value pairs further comprises:

for each of the at least two digital labels,

identifying, by the processor, one or more keys from the text object and coordinates corresponding to each of the one or more keys using the first NLP model based on aliases corresponding to the one or more attributes;

extracting, by the processor, one or more values associated with the one or more keys using the first NLP model; and

generating, by the processor, the one or more key-value pairs of attributes from the text object.

7 . A system for comparing at least two digital labels, the system comprising:

a processor; and

a memory communicably coupled to the processor, wherein the memory stores processor-executable instructions, which, on execution, cause the processor to:

detect a set of regions of interest (ROIs) corresponding to a set of objects in each of the at least two digital labels using a first Machine Learning (ML)/Deep Learning (DL) model;

classify each of the set of objects into one of a set of text objects or a set of non-text objects using a second ML/DL model;

for each text object of the set of text objects,

extract one or more key-value pairs of one or more attributes from the text object using a first Natural Language Processing (NLP) model; and

compare the one or more key-value pairs and text data corresponding to the text object of the at least two digital labels using a second NLP model;

for each non-text object of the set of non-text objects,

categorize the non-text object into an object category from a set of object categories using a third ML/DL model; and

compare an ROI of the non-text object based on positional information and the object category in the at least two digital labels; and

generate an output report based on the comparison of the text data and the comparison of the ROI of each of the set of similar type of objects in the at least two digital labels.

8 . The system of claim 7 , wherein the processor-executable instructions, on execution, cause the processor to detect a boundary of each of the at least two digital labels using an image processing technique.

9 . The system of claim 7 , wherein the set of objects comprises symbols, logos, barcodes, QR-codes, data matrix, and text data, and wherein the set of object categories comprises symbols, logos, barcodes, data matrix, and QR-codes.

10 . The system of claim 9 , wherein the processor-executable instructions, on execution, cause the processor to determine decoded data from an object in each of the at least two digital labels when the object category is one of the Barcodes, the Data Matrix, or the QR-codes.

11 . The system of claim 7 , wherein the processor-executable instructions, on execution, cause the processor to:

for each of the at least two digital labels,

identify address-associated non-text objects from the set of non-text objects;

extract address-associated text objects from the set of text objects using a Named Entity Recognition (NER) model, wherein the address-associated text objects are within a predefined threshold proximity to the address-associated non-text objects;

create masks corresponding to the address-associated text objects and address-associated non-text objects;

cluster address-associated ROIs based on the masks using image processing techniques; and

extract address information from the address-associated ROIs using a third NLP model.

12 . The system of claim 7 , wherein to extract one or more key-value pairs, the processor-executable instructions, on execution, cause the processor to:

for each of the at least two digital labels,

identify one or more keys from the text object and coordinates corresponding to each of the one or more keys using the first NLP model based on aliases corresponding to the one or more attributes;

extract one or more values associated with the one or more keys using an NLP model text analysis technique; and

generate the one or more key-value pairs of attributes from the text object.

13 . A non-transitory computer-readable medium storing computer-executable instructions for comparing at least two digital labels, the computer-executable instructions configured for:

detecting a set of regions of interest (ROIs) corresponding to a set of objects in each of the at least two digital labels using a first Machine Learning (ML)/Deep Learning (DL) model;

classifying each of the set of objects into one of a set of text objects or a set of non-text objects using a second ML/DL model;

for each text object of the set of text objects,

extracting one or more key-value pairs of one or more attributes from the text object using a first Natural Language Processing (NLP) model; and

comparing the one or more key-value pairs and text data corresponding to the text object of the at least two digital labels using a second NLP model;

for each non-text object of the set of non-text objects,

categorizing, by the processor, the non-text object into an object category from a set of object categories using a third ML/DL model; and

comparing, by the processor, an ROI of the non-text object based on positional information and the object category in the at least two digital labels; and

generating, by the processor, an output report based on the comparison of the text data and the comparison of the ROI of each of the set of similar type of objects in the at least two digital labels.

14 . The non-transitory computer-readable medium of claim 13 , the computer-executable instructions are configured for:

detecting a boundary of each of the at least two digital labels using an image processing technique.

15 . The non-transitory computer-readable medium of claim 13 , wherein the set of objects comprises symbols, logos, barcodes, QR-codes, data matrix, and text data, and wherein the set of object categories comprises symbols, logos, barcodes, data matrix, and QR-codes.

16 . The non-transitory computer-readable medium of claim 15 , the computer-executable instructions are configured for:

determining decoded data from an object in each of the at least two digital labels when the object category is one of the Barcodes, the Data Matrix, or the QR-codes.

17 . The non-transitory computer-readable medium of claim 13 , the computer-executable instructions are configured for:

for each of the at least two digital labels,

identifying address-associated non-text objects from the set of non-text objects;

extracting address-associated text objects from the set of text objects using a Named Entity Recognition (NER) model, wherein the address-associated text objects are within a predefined threshold proximity to the address-associated non-text objects;

creating masks corresponding to the address-associated text objects and address-associated non-text objects;

clustering address-associated ROIs based on the masks using image processing techniques; and

extracting address information from the address-associated ROIs using a third NLP model.

18 . The non-transitory computer-readable medium of claim 13 , the computer-executable instructions are configured for:

for each of the at least two digital labels,

identifying one or more keys from the text object and coordinates corresponding to each of the one or more keys using the first NLP model based on aliases corresponding to the one or more attributes;

extracting one or more values associated with the one or more keys using the first NLP model; and

generating the one or more key-value pairs of attributes from the text object.