研究目的
To classify the varieties of tea leaves using a multi-channel LED-induced fluorescence system combined with a CNN analytical method.
研究成果
The multi-channel LED-induced fluorescence system combined with a CNN method significantly improved the accuracy of tea classification. The system is compact, cost-effective, and robust, offering a practical approach for tea variety recognition.
研究不足
The study focused on a limited number of tea varieties and grades. The system's performance with a broader range of tea types and under varying conditions was not explored.
1：Experimental Design and Method Selection:
A multi-channel LED-induced fluorescence system was developed using seven LEDs as excitation light sources. The fluorescence spectra induced by these LEDs were analyzed using a CNN model for pattern recognition. PCA combined with kNN was used for comparison.
2：Sample Selection and Data Sources:
Nine tea samples, including six grades of green tea, two types of black tea, and one kind of white tea, were used.
3：List of Experimental Equipment and Materials:
Seven LEDs with spectra ranging from UV to blue, a spectrometer (FX2000 spectrometer; Fuxiang Inc., Shanghai, China), a microcontroller unit, and a LabVIEW program for synchronization.
4：Experimental Procedures and Operational Workflow:
LEDs were lit up sequentially to induce fluorescence spectra, which were collected by a spectrometer. The spectra were pre-processed by subtracting the background, smoothing, and normalizing.
5：Data Analysis Methods:
The pre-processed spectra were analyzed using a CNN model and compared with PCA combined with kNN.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
