Raman spectrometry has proven to be by far the most powerful noninvasive analytical technique for direct material identification. In this paper we introduce the first smart Raman device with a Cloud data platform and AI deep learning algorithms- the CloudMinds XI™. This smart phone operated Raman features high performance, fully automated operation, and capability for mixtures analysis in real time. This novel Cloud AI Raman spectrometer is fully integrated with the Android-based CloudMinds A1 smart phone. The A1 phone provides the full functionality of a smartphone including voice calls, emails, GPS location, and image capture by camera, and maintains constant Wi-Fi/blue tooth and 4G LTE connections, letting you stay connected to Raman data constantly. The cloudbased data platform not only allows speedy analysis but also enables spectral library expansion with ensured security. In addition, CloudMinds has developed its proprietary Al algorithm using Google Brain's second-generation machine learning system, TensorFlow. This technology improves analysis accuracy and gets continually better results as it learns and trains data while connected to the cloud. A mixture of three substances has been successfully analyzed with ratios within seconds by this handheld Raman spectrometer for the first time, and this paper will present the results from the mixture analysis. This Cloud AI handheld Raman is the best solution for many field applications, especially when real time analysis and central cloud data platform support are essential.
Food safety, especially edible oils, has attracted more and more attention recently. Many methods and instruments have emerged to detect the edible oils, which include oils classification and adulteration. It is well known than the adulteration is based on classification. Then, in this paper, a portable detection system, based on laser induced fluorescence, is proposed and designed to classify the various edible oils, including (olive, rapeseed, walnut, peanut, linseed, sunflower, corn oils). 532 nm laser modules are used in this equipment. Then, all the components are assembled into a module (100*100*25mm). A total of 700 sets of fluorescence data (100 sets of each type oil) are collected. In order to classify different edible oils, principle components analysis and support vector machine have been employed in the data analysis. The training set consisted of 560 sets of data (80 sets of each oil) and the test set consisted of 140 sets of data (20 sets of each oil). The recognition rate is up to 99%, which demonstrates the reliability of this potable system. With nonintrusive and no sample preparation characteristic, the potable system can be effectively applied for food detection.
In order to accomplish recognition of the different edible oil we set up a laser induced fluorescence spectrum system in the laboratory based on Laser induced fluorescence spectrum technology, and then collect the fluorescence spectrum of different edible oil by using that system. Based on this, we set up a fluorescence spectrum database of different cooking oil. It is clear that there are three main peak position of different edible oil from fluorescence spectrum chart. Although the peak positions of all cooking oil were almost the same, the relative intensity of different edible oils was totally different. So it could easily accomplish that oil recognition could take advantage of the difference of relative intensity. Feature invariants were extracted from the spectrum data, which were chosen from the fluorescence spectrum database randomly, before distinguishing different cooking oil. Then back propagation (BP) neural network was established and trained by the chosen data from the spectrum database. On that basis real experiment data was identified by BP neural network. It was found that the overall recognition rate could reach as high as 83.2%. Experiments showed that the laser induced fluorescence spectrum of different cooking oil was very different from each other, which could be used to accomplish the oil recognition. Laser induced fluorescence spectrum technology, combined BP neural network,was fast, high sensitivity, non-contact, and high recognition rate. It could become a new technique to accomplish the edible oil recognition and quality detection.
The pure rotational Raman lidar temperature measurement system is usually used for retrieval of atmospheric temperature according to the echo signal ratio of high and low-level quantum numbers of N2 molecules which are consistent with the exponential relationship. An effective method to detect the rotational Raman spectrum is taking a double grating monochromator. In this paper the detection principle and the structure of the dual-grating monochromator are described, with analysis of rotational Raman’s Stokes and anti-Stokes spectrums of N2 molecule, the high order and lower order quantum number of the probe spectrum are resolved, then the specific design parameters are presented. Subsequently spectral effect is simulated with Zemax software. The simulation result indicates that under the condition of the probe laser wavelength of 532nm and using double-grating spectrometer which is comprised by two blazed gratings, Raman spectrums of 529.05nm, 530.40nm, 533.77nm, 535.13nm can be separated well, and double-grating monochromator has high diffraction efficiency.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.