ENOVISION

The ENOVISION system works on a sensor fusion technology where an array of non-specific Metal Oxide Semiconductor (MOS) Sensors captures the volatile compounds emanated by Black Tea samples and a digital camera captures the image of the sample under a uniform illumination condition. The multi-dimensional MOS sensor outputs along with the captured image by the digital camera are stored in the PC through PC-based Data Acquisition Systems. These stored image and aroma signatures of the Black Tea sample are analyzed through intelligent pattern recognition algorithm for clustering, classification and measurement of Black tea quality in terms of aroma and appearance.

The hardware part of the system consists of a) odour handling and delivery system, b) sensor array PCB, c) data acquisition PCB with capability of analog and digital I/O and USB interface, d) valve control PCB, e) solenoid valves for headspace control and f) enclosure cabinet of the system g) digital camera h) Uniform LED based illumination i) Sample holding tray with different types of sample holding attachments.

The software is based on Windows® platform. The E-Nose software is developed with National Instrument LabVIEW® and the E-Vision software is developed using Visual C++. Microsoft Access forms the backend database for the software.

Electronic Nose:

Electronic Nose is a smart instrument that is designed to detect and discriminate amongst complex odours using an array of sensors. The array of sensors is made up of a number of broadly tuned and non-specific sensors.

An odour stimulus generates a characteristic fingerprint from this array of sensors. Patterns or fingerprints from known odours are used to construct a database and train a pattern recognition algorithm so that unknown odours can subsequently be classified and/or identified.

Such an Electronic Nose system consists of four functional blocks, namely: (a) Odour Handling and Delivery System, (b) Sensors and Interface Electronics, (c) Signal Processing and (d) Intelligent Pattern Analysis and Recognition. The array of sensors is exposed to volatile odour vapour through suitable odour handling and delivery system that ensures constant exposure rate to each of the sensors. The response signals of sensor array are conditioned and processed through suitable circuitry and fed to an intelligent Pattern Recognition Engine for classification, analysis and measurement.

The most complicated parts of electronic olfaction process are odour capture and associated sensor technology. Any sensor that responds reversibly to chemicals in gas or vapour phase, has the potential to participate as a component of an array of sensor in an electronic nose. For black manufactured tea, an array of Metal Oxide Semiconductor (MoS) sensors has been used for measurement of volatiles.

Application of Electronic Nose

• E-Nose for Fermentation Monitoring
• E-Nose for Finished Tea Classification

Electronic Vision:

Electronic Vision system is a non destructive, non-invasive, fast, quantifiable (objective) quality monitoring instrument of food and agro produces based on the appearance features of the product. It takes colour images of the sample in a constant illumination and uses digital image processing cum analysis techniques to find out different features of the object and finally determine the quality of the target object based on previously stored data which was taken during training of the system or by statistical algorithms. A number of software applications have been developed for different type of applications.

It uses a low cost digital camera which has Universal Serial Bus Interface with computer, white colored light emitting diode (LED) based illumination system which consume low power, rail mounted movable tray based easy to use sample holding arrangement with numerous attachment for different type of sample (liquid sample like tea liquor, solid sample like finished tea etc) holding attachments. To avoid the external interference of the light the whole system was designed as a enclosed cabinet, inside which the camera, illumination setup, sample holding tray have been placed. The digital camera takes the colour image, image processing algorithm process the image to highlight the specific features by enhancing the image contrast, removing the noise, segment the image from the background etc. Different types of image analysis algorithms (e.g. colour analysis, particle size analysis, shape detection, morphological features (area, perimeter, ratio of length and breadth, compact factor, shape factor, roundness etc) have been used to analysis the images based on the specific applications.

Applications of E-Vision System for Tea

• To detect the end point of fermentation
• Applications for estimation of percentage of different tea grades in bulk tea
• Quality testing of Tea based on the appearance property 

Use Cases

• Regular assessment of tea quality during tea production by the Tea Industry.
• Quality testing of tea sample during Tea Auction.
• Testing of quality of tea samples during tea blending.

Salient Features

E-Nose Feature:

• E-Nose can dynamically indicate olfactory emission pattern in
• Fermentation.
• Online display of fermentation progress and completion.
• Accurate detection of 1st and 2nd Nose Peaks during fermentation.
• Reliable Prediction of Tea Taster-like Quality Scores of Finished Tea.
• Easy to use.
• Instant data acquisition.
• Zero sample preparation.

E-Vision Feature:

• GUI with automatic/manual operation.
• Online/Offline analysis.
• Instant Report Generation/Storing/e-mail.
• Visual/Audio Alarm.
• Fermentation colour profile display/ comparison.
• Flexibility for customizing the S/W for Tea Garden/Taster & Tea Taster-like Quality Index.