Data Processing Of Portable Raman Spectrometer And System Design For Anti-counterfeit

The fake and shoddy products is the world’s second largest public nuisance after drug trafficking. Traditional Ink anti-counterfeiting technology such as fluorescent falsification-resistant ink, heat sensitive ink and pressure sensitive ink has been used for many years and lost anti-counterfeiting function by and large. As reflecting the molecular structure and component information, Raman spectrum, a kind of molecule scattering spectroscopy, has fingerprint identification function of material. It is hard to imitate when ordinary nano materials can be combined into new materials with specific Raman spectrum by professional means. There are many advantages with the Raman spectrum analysis technology, such as non-destructive, highly sensitive, non-preparative sample and short time to test. As a consequence, based on analysis of Raman spectra, the new anti-counterfeiting technology has more advantages than traditional ink by adding nano materials into ink.We have exploratory research in the field of barcode security. Based on the first portable Raman spectrometer prototype of self-assembly, the data processing of Raman spectrum automatically is studied systematically in this paper, which is the basis of the qualitative analysis of smart nanobarcodes encoding technology and the analysis software designed to be put into use. The major research work is as follows:1. Studied the principle of nanobarcodes encoding based on Raman spectrum of nano materials, by changing the combination to realize the barcode encoding and expand the information capacity. With simple encoding and large information capacity, it can meet the needs of the actual commodity barcode encoding. Besides, briefly summarize the overall framework of the subject and the hardware structure and characteristics of the portable Raman spectrometer prototype of self-assembly.2. Studied the data preprocessing methods and the peak recognition of Raman spectrum, including the denoise and debaseline methods. Realized an automated noise reduction method based on three points zero-order Savitzky-Golay filter. Compared with four kinds of traditional denoise methods, numerical experiments showed that the method had a good denoising performance. It could retain most of detail information contained in Raman peaks. Realized a baseline estimation method, small window moving average method, which is a totally automatic debaseline method without any parameters to be set. Compared with three kinds of commonly used methods, it showed that the best baseline estimation accuracy of those methods was the one we use. Realized the peak recognition method based on bi-scale correlation. Simulation results showed that the bi-scale correlation method outperformed other traditional methods by compared with peak recognition performance of them. It has no needs of de-noising and background removed operation to avoid error caused in those process. Analysis showed that no matter to singular peak or congested peaks, when the signal-to-noise ratio of Raman peak is greater than or equal to 8, the recognition accuracy of bi-scale correlation algorithm is higher than 99.8%.3. Developed a smart recognition system to detect nano materials that barcode contains with friendly interface based on the core algorithm we studied above. Refine the technology architecture of the system from system requirements. After analyzed the overall system framework, function modules and core algorithm implementation process, we designed of the database and completed code. Finally, functional test is completed to meet the requirements of practical application.