Tobacco sorting system with image processing and IoT

Abstract

A system for the automated evaluation of tobacco leaves, comprising: • a mechanical feeding module (201) configured to feed individual tobacco leaves one after the other; • a controlled imaging chamber (202) equipped with a high-resolution camera and calibrated light sources to capture detailed images of tobacco leaves under standardized lighting conditions; • a color sensor (203) configured to extract colorimetric data from the tobacco leaves, including, but not limited to, RGB values and hue information; • a processing unit (204) comprising one or more processors and memories configured to: i) performing image pre-processing including noise reduction and contrast enhancement, (ii) Execution of segmentation algorithms to isolate the leaf from the background, (iii) Extraction of visual features such as size, shape, texture and color uniformity from the images, (iv) Analyzing the sensor data using statistical or machine learning models to determine quality parameters, (v) Integration of the extracted image features and colorimetric data into an evaluation algorithm based on predefined quality standards, (vi) to assign a quality mark to each tobacco leaf on the basis of the analysis; • an output module (205) configured to display the classification result in real time via an LED screen or a digital interface; • a storage module (206) configured to store classification results for future reference and analysis; • and an Internet of Things (IoT) communication module (207) that enables remote access, monitoring and control of the classification process.

Description

Field of the invention:

[0001] The present invention relates to the field of agricultural engineering. More specifically, it relates to an automatic tobacco sorting system that uses Internet of Things (IoT)-based sensors and image processing techniques to evaluate the quality of tobacco leaves. The invention addresses the shortcomings of conventional manual classification methods and aims to improve accuracy, consistency, and operational productivity in the tobacco industry. Background of the invention:

[0002] Tobacco is one of the most economically important crops, grown in many parts of the world and used as a raw material for various products such as cigarettes, cigars, and chewing tobacco. The quality of tobacco leaves plays a crucial role in determining their suitability for these uses and significantly impacts their market value. Tobacco grading is an essential post-harvest process in which leaves are evaluated and classified based on various quality characteristics such as color, size, texture, maturity, and physical defects. Traditionally, this process is labor-intensive and requires skilled personnel to manually inspect the leaves. Such manual grading is inherently subjective, highly variable, time-consuming, and costly. Furthermore, the process can be strenuous and error-prone, especially when large quantities of leaves need to be evaluated within a limited timeframe.

[0003] In traditional operations, sorting tobacco leaves can take several weeks to months, depending on the scope and complexity of the sorting standards to be met. Grading consistency is often questionable, as assessment by different employees, or even by the same employee, varies over time. Furthermore, the manual approach is resource-intensive and requires trained personnel and infrastructure, significantly increasing operating costs. In many regions, particularly where there is an acute labor shortage or skilled sorters are not readily available, tobacco growers struggle to effectively meet sorting requirements, leading to potential loss of market value or product rejection.

[0004] As global demand for quality-assured agricultural products increases, there is an urgent need to modernize and automate traditional methods of crop evaluation. In recent years, the integration of advanced technologies into agricultural practices has gained momentum. Precision agriculture, which combines sensors, data analytics, and automation, has transformed the way farmers monitor and manage the health, yield, and quality of their crops. Despite these advances, post-harvest processing systems, such as tobacco grading, have not been significantly modernized and still largely rely on manual methods.

[0005] One of the most promising solutions to address this challenge lies in the fusion of machine vision and the Internet of Things (IoT). Machine vision enables machines to mimic human visual perception, enabling automated assessment of leaf physical characteristics such as size, color, and surface texture. IoT technologies facilitate the collection, processing, and transmission of data in real time through interconnected sensors and devices. Together, these technologies can form the basis of an intelligent classification system that enables objective, consistent, and rapid classification of tobacco leaves.

[0006] Several efforts have been made to apply machine vision to agricultural applications such as fruit ripeness detection, disease identification, and quality control in product packaging. However, the application of these technologies to tobacco grading remains underexplored, particularly in developing regions where tobacco cultivation is widespread. Furthermore, an effective tobacco grading system must go beyond mere image capture; it should include systematic processing of visual and colorimetric data, reliable classification algorithms, and an integrated feedback mechanism to inform users in real time. Furthermore, data management capabilities are critical for quality traceability and for integrating the system into broader agricultural supply chains.

[0007] The need for such an automated classification system is further compounded by the challenges faced in regions with limited access to skilled labor or where producers must adhere to strict quality regulations to access export markets. A reliable, sensor-based, image-guided classification system can help tobacco producers comply with international quality standards, reduce operating costs, and increase the overall profitability of tobacco production. By adopting such technological solutions, the industry can ensure a more transparent, efficient, and scalable classification mechanism that benefits all stakeholders along the value chain—from farmers and processors to traders and end consumers. Summary of the invention:

[0008] The present invention introduces an intelligent and automated tobacco grading system that utilizes image processing and Internet of Things (IoT) technologies to classify and grade tobacco leaves with greater efficiency, accuracy, and consistency than traditional manual methods. This innovation addresses the long-standing challenges associated with conventional tobacco grading, particularly subjectivity, labor dependence, and operational delays, which often compromise quality control and increase production costs.

[0009] At the core of the system is a well-integrated combination of hardware and software modules that cover the entire sorting process. Tobacco leaves are first fed into a leaf unit, which ensures that each leaf is processed individually to maintain uniformity. As the leaf enters the sorting module, a high-resolution digital camera captures detailed images, while a color sensor measures color characteristics such as tone, intensity, and hue. These two data streams—image data and color sensor data—are then processed by an onboard computer unit equipped with image processing algorithms designed to evaluate characteristics such as the leaf's shape, texture, and ripeness.

[0010] The analyzed data is then compared against predefined evaluation criteria based on industry standards and expert opinion. Each leaf receives a grade reflecting its overall quality, with the listing decision displayed in real time on an LED screen or connected computer interface. The system also supports database features that allow historical tracking of evaluation results and facilitate further analysis or reporting. This makes the system not only an evaluation tool but also a data management solution that can be integrated into broader digital agricultural ecosystems.

[0011] One of the main advantages of the invention is its adaptability. The grading parameters can be adjusted according to the specific requirements of different regions, tobacco varieties, or end-user preferences. The system is designed to be scalable and portable, enabling its use in both large-scale industrial settings and smaller farm-level operations. By reducing dependence on skilled labor, the system enables producers to perform accurate and consistent grading even in areas where skilled labor is scarce.

[0012] Furthermore, the system's real-time processing and feedback capabilities drastically reduce processing time for grading large volumes of leaves. This not only speeds up the supply chain but also minimizes delays in packaging and shipping. The consistency provided by machine sorting also supports fair pricing mechanisms by eliminating human bias and variability in quality assessment.

[0013] The invention also emphasizes user-friendliness. The user interface is designed to be intuitive, allowing even non-technical users to operate the system with minimal training. Maintenance requirements are minimal, and the system can be integrated into existing agricultural management platforms to further increase operational efficiency.

[0014] In conclusion, the proposed tobacco sorting system offers a transformative solution to one of the most significant bottlenecks in tobacco processing. By automating the sorting process with IoT and machine vision technologies, the invention ensures higher accuracy, faster processing, and reduced labor costs. This innovation has the potential to raise quality standards, improve profitability, and set new standards in agricultural technology, particularly in post-harvest tobacco processing. Short description of the drawing Fig. shows a block diagram of the system according to the invention. Detailed description of the invention

[0015] The present invention relates to a comprehensive system and method for automated tobacco sorting that utilizes image processing techniques and the integration of Internet of Things (IoT)-based sensors to enable real-time, accurate, and efficient classification of tobacco leaves. This system is specifically designed to overcome the traditional limitations of manual sorting by providing consistent quality assessments, minimizing human error, and optimizing post-harvest operations in the tobacco industry.

[0016] The invention begins with the construction of a mechanical feeding system that picks up individual tobacco leaves in a sequential manner to ensure consistent processing. The tobacco leaves fed into the system pass through a controlled environment that isolates external light disturbances, thus enabling consistent imaging conditions. This chamber is equipped with a high-resolution camera and a series of lighting elements calibrated to produce standardized illumination, which is critical for reliable image acquisition. The camera is capable of capturing fine details of each leaf, including its edges, texture patterns, veining, and surface anomalies. At the same time, an advanced color sensor is employed to extract colorimetric data, including RGB values and tonal characteristics, which are critical for determining leaf maturity and quality.

[0017] After capturing the raw image and sensor data, the system moves to the central processing unit, where image preprocessing steps are performed. These include noise reduction, contrast enhancement, and image normalization to eliminate inconsistencies due to environmental conditions or sensor variations. The image is then subjected to segmentation algorithms that distinguish the leaf from the background. This is followed by feature extraction processes that quantify key visual parameters such as size, shape, color uniformity, and surface texture. The extracted features are compared against predefined scoring guidelines derived from agricultural standards and expert knowledge, which are stored in the system's scoring database.

[0018] The color sensor plays a complementary role in the sorting process by providing an additional layer of data. While image processing captures visual structure and texture, the color sensor provides precise measurements of leaf coloration, often a critical quality indicator. The sensor data is processed using statistical and machine learning models trained on a large dataset of annotated leaf samples. These models can distinguish between subtle color variations that are not easily perceptible to the naked eye, thereby improving classification accuracy.

[0019] The integrated processing framework combines both image-based features and sensor data into a multi-layered classification model. This model uses decision trees and neural network algorithms to assign a quality label to each leaf. The system supports multi-grade categorization, which allows the classification of leaves into different commercial grades such as premium, standard, and substandard, or into specific industry-defined grades. The classification result is visually displayed on an LED display panel and also digitally logged for archiving and analysis purposes.

[0020] To support scalability and adaptability, the system includes a modular software interface that allows users to update evaluation parameters, retrain models, and review historical data. This interface is accessible via a connected computer or mobile device and provides real-time monitoring and remote control capabilities. Additionally, the system architecture supports integration with cloud platforms for data synchronization, centralized analytics, and long-term storage. These cloud-based capabilities facilitate the creation of a digital inventory of sorted tobacco, contributing to supply chain transparency and traceability.

[0021] Furthermore, the system's integrated IoT connectivity ensures that all sensors, imaging components, and output modules are connected via a central controller. This controller continuously monitors the operating status of each component, flags malfunctions, and dynamically adjusts processing parameters based on feedback. For example, if the lighting conditions in the imaging room deviate due to aging LEDs, the system compensates by recalibrating the camera exposure or issuing maintenance alerts.

[0022] The invention also incorporates safety and operational safeguards, such as automated shutdown protocols in the event of system failures, emergency stop mechanisms for the feed module, and a protective housing for all electronic components to withstand agricultural dust and moisture. These design considerations ensure the robustness and reliability of the system in field environments.

[0023] In its preferred embodiment, the system is housed in a compact, mobile-friendly unit that can be deployed in collection centers, auction houses, or directly on farms. Mobility is facilitated by a robust wheeled chassis, onboard power options including solar compatibility, and wireless communication modules for real-time data transmission. This design makes the system particularly suitable for deployment in rural or resource-poor areas where traditional valuation infrastructure may be lacking.

[0024] The real-time rating output is not only displayed for immediate viewing but also stored in a structured database that can be queried for batch analysis. Reports can be automatically generated at predefined intervals, summarizing the distribution of ratings, average quality scores, and historical trends. These reports are invaluable for farmers, traders, and quality assurance personnel who need to make informed decisions about pricing, storage, and logistics.

[0025] To further enhance user-friendliness, the invention features an intuitive user interface available in multiple languages and adaptable to specific regional requirements. This interface provides step-by-step guidance on operating the system, viewing assessment results, and performing routine maintenance. The system also features plug-and-play capabilities, allowing for easy replacement of sensors or cameras without complex recalibrations.

[0026] The applicability of the invention extends beyond tobacco to include other leafy agricultural products where color, texture, and size are critical evaluation factors. Therefore, the modular architecture and learning algorithms can be retrained for alternative use cases, expanding the potential market for the technology.

[0027] In summary, the present invention provides a technologically advanced, reliable, and user-friendly solution for evaluating tobacco leaves. Combining high-resolution imaging, precise color recognition, intelligent data processing, and IoT-enabled automation, the system transforms the evaluation process from a subjective, labor-intensive task into a standardized, data-driven operation. The invention not only eliminates the inefficiencies of manual sorting but also introduces new opportunities for data analysis, traceability, and operational integration within the agricultural value chain. This leads to improved quality assurance, reduced post-harvest losses, and increased economic returns for all stakeholders in tobacco production and processing. List of reference symbols 201 Mechanical feeding module 202 Controlled imaging chamber 203 color sensor 204 processing unit 205 Output module 206 memory module 207 Internet of Things (IoT)

Claims

[1] A system for the automated evaluation of tobacco leaves, comprising: • a mechanical feeding module (201) configured to feed individual tobacco leaves one after the other; • a controlled imaging chamber (202) equipped with a high-resolution camera and calibrated light sources to capture detailed images of tobacco leaves under standardized lighting conditions; • a color sensor (203) configured to extract colorimetric data from the tobacco leaves, including, but not limited to, RGB values and hue information; • a processing unit (204) comprising one or more processors and memories configured to: i) performing image pre-processing including noise reduction and contrast enhancement, (ii) Execution of segmentation algorithms to isolate the leaf from the background, (iii) Extraction of visual features such as size, shape, texture and color uniformity from the images, (iv) Analyzing the sensor data using statistical or machine learning models to determine quality parameters, (v) Integration of the extracted image features and colorimetric data into an evaluation algorithm based on predefined quality standards, (vi) to assign a quality mark to each tobacco leaf on the basis of the analysis; • an output module (205) configured to display the classification result in real time via an LED screen or a digital interface; • a storage module (206) configured to store classification results for future reference and analysis; • and an Internet of Things (IoT) communication module (207) that enables remote access, monitoring and control of the classification process. [2] The system of claim 1, wherein the processing unit is further configured to use a neural network-based classification model trained on annotated tobacco leaf datasets to assign quality marks. [3] The system of claim 1, wherein the color sensor is capable of detecting color attributes that are not easily perceptible to the human eye, thereby improving the granularity and accuracy of the classification process. [4] The system of claim 1, wherein the IoT communication module is integrated into a cloud platform to enable centralized storage, real-time synchronization of sorting data, and the generation of analytical reports. [5] The system of claim 1, wherein the user interface supports multilingual functionality and provides interactive guidance for system operation, maintenance, and interpretation of classification results. [6] The system of claim 1, wherein the mechanical feed module includes a sensor-based alignment mechanism to ensure consistent alignment and positioning of each sheet for imaging and analysis.

Images