Spectroscopy and Machine Vision Solutions
NIR spectroscopy, or Near Infrared Spectroscopy, is an analytical technique for determining the chemical composition and certain physical properties of various materials and products based on the analysis of the interaction of optical radiation (light) with the molecular and atomic structures of these materials. For this reason, NIR is a very widespread technique in the analysis of food, grains, mixtures and pharmaceutical products, chemicals, cosmetics and even in other industries such as plastics for the identification of polymers, recycling, wood, among others.
In practical terms, NIR spectroscopy involves the development of machine learning models that convert spectral information into analytical parameter values, whether quantitative or qualitative.
Spectroscopy -and especially NIR spectroscopy– is becoming a versatile and robust tool in what is known as PAT (process analytical technologies), that is, the control of production processes and the quality of the finished product based on knowing, in real time and online, the composition and physical properties of what is intended to be produced and not only having information on process conditions.
For this reason, its applicability ranges from the exact determination of water content to the determination of the final mixing point, not to mention the degree of curing (polymerization), the identification of materials for separation, the validation of raw materials or even the detection of anomalies due to accidents or fraudulent practices.
The obvious advantage is the time needed to get the result and thus to make a decision on the ongoing process. With a spectroscopic probe the result is available in fractions of a second, whereas laboratory results can take from hours to days.
Also, as the determination is fully automatic, distorting factors in traditional analysis inherent in human intervention, both in sampling and analysis, are eliminated.
Inline NIR spectroscopy is generally not suitable for the determination of concentrations below 0.1% by weight, with 0.5% by weight being a fairly common threshold for applications of this technology when used inline or in real-time.
Click here to view real-time solutions and Visum® NIR spectrometers.
Unlike NIR spectroscopy, where water can be a considerable interfering factor, water is not a problem for the Raman signal. Both spectroscopies can be complementary in terms of possible applications: given a certain practical case, the results achievable by both technologies have to be compared in order to select the most suitable one, as the factors to be considered are not only metrological.
However, a number of applications of Raman spectroscopy are listed below:
- Characterisation of crystallisation processes.
- Control of the polymerisation process.
- Identification of components.
- Hydrogenation reactions.
- Quantification of active pharmaceutical ingredients (APIs) in low concentrations.
- Control of the fermentation process and extraction of APIs.
- Characterisation of organic or inorganic substances.
- Geology and mineralogy.
- Materials research.
Like NIR spectroscopy, Raman probes are ideal media for use as PAT tools. Unlike NIR spectroscopy, where water can be a considerable interfering factor, water is not a problem for the Raman signal. The two spectroscopies can be complementary in terms of possible applications: given a certain case study, the results achievable by both technologies have to be compared in order to select the most suitable one, as the factors to be considered are not only metrological.
Artificial vision based on deep learning (neural networks) mimics, to a certain extent, the way in which the human brain processes visual information, so that, as with the brain, it is essential that there is first a learning process. To do this, a sufficiently representative set of images of the cases to be detected must be available, which serves to “train” the neural network, i.e. the network must be able to “interpret” or “recognise” the characteristics of interest. Once the constructed network has been validated, it is used to inspect each new image, the output of the network being a label that identifies that image as belonging to one of the pre-established classes. In practice, a neural network is a set of numerical values, collected in a digital file, specific to a particular task, so that dedicated software “applies” the network to each image in order to obtain the label corresponding to the class to which the image appears to belong.
Los algoritmos de visión artificial tradicionales son determinísticos. Eso significa que, mediante funciones matemáticas, sean genéricas o específicas, se analiza cada imagen buscando las características de interés. Ese enfoque limita su aplicación a casos muy concretos, en los que las características sean muy marcadas o evidentes respecto al resto de los elementos de la imagen y siempre requiere de la intervención de expertos humanos. En el mundo real las circunstancias favorables al uso de algoritmos determinísticos no son nada frecuentes. Por el contrario, hay muchos elementos interferentes que afectan sobremanera la exactitud del resultado. El deep learning, por el contrario, no parte de supuestos más o menos realistas, sino que, en la fase de aprendizaje, se adapta automáticamente a las circunstancias, esto es, es menos sensible a variaciones imprevistas, y ello puede conseguirlo sin necesidad de supuestos teóricos o del trabajo de un experto.
The most frequent application is the detection of morphological defects in order to activate a rejection system that discards those products labelled by the neural network as defective. For example, detection of foreign bodies or defective units in foodstuffs, intolerable colour differences in parts, textiles or food products or the objective quantification of faults or defects in order to calculate a fair price for the various quality grades of a product, for example defects in grains for sale and marketing.