US 2026/0301191 A1Published application

Method and system for realtime automated grading of tea shoots for quality and resource management

AssigneeTata Consultancy Services Limited

InventorsSANKET KAILAS JUNAGADE, SWAGATAM BOSE CHOUDHURY, SANAT SARANGI, DINESHKUMAR JANG BAHADUR SINGH, SRINIVASU PAPPULA

Life of this patent

Estimating the term…

patentgrep

Loading patent…

Abstract

A method and system for real time automated grading of tea shoots for quality and resource management is disclosed. The method disclosed implements a multi-layer data fusion approach integrating RGB and spectral data of tea leaves for simultaneous morphological and chemical profiling. The method provides real-time detection of critical anomalies and an anomaly scoring mechanism. The method links real-time chemical profiling data with flavor prediction models and adjusts processing parameters based on target flavor profile. The method provides a dynamic worker compensation mechanism which includes real-time compensation to leaf quality metrics, incorporating post-segregation anomaly handling and rewarding performance based on output quality. The method allocates rejected leaves to optimized repurpose methods, minimizing waste and promoting sustainability. Thus, the method disclosed enables real-time tea leaf classification, flavor profile prediction, manufacturing optimization, dynamic worker compensation, waste management, and comprehensive reporting in tea manufacturing industry.

Claims20

Dependent claims

A processor implemented method comprising:

segmenting, via one or more hardware processors, a set of RGB images and a set of spectral images corresponding to a batch of tea leaves, captured via an RGB camera and a spectral camera for a pre-defined time interval, by applying
(i) a first segmentation technique on the set of RGB images to obtain a set of segmented RGB images, and
(ii) at least (a) a second segmentation technique on a set of histogram-equalized spectral images to obtain a first subset of segmented spectral images amongst a set of segmented spectral images and (b) a third segmentation technique on the set of histogram-equalized spectral images to obtain a second subset of segmented spectral images amongst the set of segmented spectral images;
extracting, via the one or more hardware processors, a set of features comprising (i) a set of morphological features by processing the set of segmented RGB images, and (ii) a set of spectral features by applying a spectral reflectance extraction technique and a dimensionality reduction technique on the set of segmented spectral images;
selecting, via the one or more hardware processors, one or more feasible grades amongst a set of predefined grades based on a feasibility score calculated using (i) the set of features and (ii) a set of shoot classes relevant to tea manufacturing;
determining, via the one or more hardware processors, a flavor score for the one or more feasible grades based on (i) a set of flavor compounds, and (ii) a bitterness index calculated using one or more flavor compounds amongst the set of flavor compounds;
validating, via the one or more hardware processors, one or more final tea grades for a batch of processed tea, based on a Reinforcement Learning (RL) technique with a reward function, wherein the reward function is based on the flavor score, a user-defined target flavor profile, and a user defined grade;
and
calculating, via the one or more hardware processors, a total worker compensation by correlating the set of features, a set of anomalies comprising (i) a set of pre-informed anomalies, and (ii) a set of uninformed anomalies, detected from the set of features, and a set of scoring rules.

6 more elements

A system comprising:

a memory storing instructions;
one or more communication interfaces;
and
one or more hardware processors coupled to the memory via the one or more communication interfaces, wherein the one or more hardware processors are configured by the instructions to:
segment a set of RGB images and a set of spectral images corresponding to a batch of tea leaves, captured via an RGB camera and a spectral camera for a pre-defined time interval, by applying
(i) a first segmentation technique on the set of RGB images to obtain a set of segmented RGB images, and
(ii) at least (a) a second segmentation technique on a set of histogram-equalized spectral images to obtain a first subset of segmented spectral images amongst a set of segmented spectral images and (b) a third segmentation technique on the set of histogram-equalized spectral images to obtain a second subset of segmented spectral images amongst the set of segmented spectral images;
extract a set of features comprising (i) a set of morphological features by processing the set of segmented RGB images, and (ii) a set of spectral features by applying a spectral reflectance extraction technique and a dimensionality reduction technique on the set of segmented spectral images;
select one or more feasible grades amongst a set of predefined grades based on a feasibility score calculated using (i) the set of features and (ii) a set of shoot classes relevant to tea manufacturing;
determine a flavor score for the one or more feasible grades based on (i) a set of flavor compounds, and (ii) a bitterness index calculated using one or more flavor compounds amongst the set of flavor compounds;
validate one or more final tea grades for a batch of processed tea, based on a Reinforcement Learning (RL) technique with a reward function, wherein the reward function is based on the flavor score, a user-defined target flavor profile, and a user defined grade;
and
calculate a total worker compensation by correlating the set of features, a set of anomalies comprising (i) a set of pre-informed anomalies, and (ii) a set of uninformed anomalies, detected from the set of features, and a set of scoring rules.

10 more elements

One or more non-transitory machine-readable information storage mediums comprising one or more instructions which when executed by one or more hardware processors cause:

segmenting, a set of RGB images and a set of spectral images corresponding to a batch of tea leaves, captured via an RGB camera and a spectral camera for a pre-defined time interval, by applying
(iii) a first segmentation technique on the set of RGB images to obtain a set of segmented RGB images, and
(iv) at least (a) a second segmentation technique on a set of histogram-equalized spectral images to obtain a first subset of segmented spectral images amongst a set of segmented spectral images and (b) a third segmentation technique on the set of histogram-equalized spectral images to obtain a second subset of segmented spectral images amongst the set of segmented spectral images;
extracting a set of features further comprising (i) a set of morphological features by processing the set of segmented RGB images, and (ii) a set of spectral features by applying a spectral reflectance extraction technique and a dimensionality reduction technique on the set of segmented spectral images;
selecting one or more feasible grades amongst a set of predefined grades based on a feasibility score calculated using (i) the set of features and (ii) a set of shoot classes relevant to tea manufacturing;
determining a flavor score for the one or more feasible grades based on (i) a set of flavor compounds, and (ii) a bitterness index calculated using one or more flavor compounds amongst the set of flavor compounds;
validating one or more final tea grades for a batch of processed tea, based on a Reinforcement Learning (RL) technique with a reward function, wherein the reward function is based on the flavor score, a user-defined target flavor profile, and a user defined grade;
and
calculating a total worker compensation by correlating the set of features, a set of anomalies further comprising (i) a set of pre-informed anomalies, and (ii) a set of uninformed anomalies, detected from the set of features, and a set of scoring rules.

6 more elements

Dependencies are inferred from the claim text.