Multispectral Imaging Food Detection Device Based on Narrowband Absorber in Visible Light Band

Abstract

A multispectral imaging food detection device based on a narrowband absorber in the visible light band belongs to the field of spectral imaging food detection. The present invention designs a multispectral imaging system based on a visible light narrowband absorber, arranges the designed absorbers into a two-dimensional array as the imaging unit of the detector to directly form a visual image. The microstructure of the absorber adopts a sandwich structure waveguide formed by metal-dielectric-metal. The top and bottom metal layers can be directly used as electrodes and integrated into the electro-optical system, miniaturizing the multispectral imaging system; because the absorption peak of the narrowband absorber designed in the present invention is continuously adjustable, it can be applied to the multispectral imaging of different fruits, vegetables and root vegetables.

Description

Technical Field

[0001] The present invention belongs to the field of spectral imaging food detection, and particularly relates to a multispectral imaging food detection device based on a narrowband absorber in the visible light band. Background Art

[0002] Food safety issues are related to human health and life safety and have long been a hot topic for scientific research workers. Multispectral imaging technology can capture image data in discontinuous spectral ranges, and this technology is now considered an advanced means to meet the safety detection needs of the food industry and has been widely used in recent years for the prediction and visualization of various quality parameters of plant-based foods.

[0003] Plant-based foods play a central role in meeting the growing global demand for agricultural products. As an extremely important part of a healthy diet, plant-based foods not only provide a large amount of energy and nutrients but also reduce the risks to humans from many diseases (such as insulin resistance, diabetes, and heart disease). Food safety issues must be effectively monitored by relevant departments in the food industry. Any pathogenic, adulterated, contaminated, or defective plant-based foods, whether intentional or unintentional, will not only reduce the appearance characteristics and nutritional value of plant-based foods but may also cause a series of foodborne diseases. Therefore, it is very important and necessary to evaluate the quality of plant-based foods. It is worth mentioning that hyperspectral imaging technology can integrate the spectral and imaging characteristics into a system to provide heterogeneous information reflecting food quality characteristics and has been widely studied in recent years. A hyperspectral image refers to an image in which each pixel almost forms a continuous spectrum. However, the development of real-time detection technology based on hyperspectral imaging has encountered bottlenecks because the speed of acquiring and processing hundreds of continuous narrowband images in hyperspectral sensors is limited. Multispectral imaging processes narrowband images in discrete spectral ranges and can generate characteristic wavelengths for each pixel in the target. Given that multispectral imaging can capture image data in discontinuous spectral ranges, this technology is now considered an advanced means to meet the speed requirements of the food industry. Since multispectral images can convert the acquired three-dimensional data into two-dimensional chemical images through multivariate modeling, multispectral imaging technology has been widely used in recent years for the prediction and visualization of various quality parameters of plant-based foods. Such as physical properties (water-holding capacity), chemical aspects (such as starch, protein), microbial aspects (such as viruses, fungi), adulteration (such as species, origin), grading (such as maturity period, storage period), contamination (such as imidacloprid, clothianidin), and defects (such as bruising, rot).

[0004] Multispectral imaging systems typically use multiple narrow-band filters to preferentially transmit specific spectral components to different detectors, thereby giving images generally in 3 to 20 bands. They are typically characterized by a set of detectors encoded with unique spectral response characteristics, which can be combined using complex algorithms when measured in parallel to approximate or "reconstruct" the incident spectrum. However, traditional multispectral imaging systems rely on a combination of bulky dispersive optics, long optical path lengths, detector arrays, and movable components. These requirements hinder the miniaturization of applications. In recent years, perfect absorbers designed and fabricated using the unique properties of metamaterials in multispectral imaging have become one of the research hotspots. Multi-resonance usually mixes resonators of different sizes so that they each absorb the corresponding wavelength, and the unit period is less than the incident wavelength. Therefore, it is mostly applied in multi-band absorbers in the microwave, terahertz, and infrared bands. When the resonant wavelength is extended to the visible band, since the structure period itself is less than the resonant wavelength, the size of a single resonant element within the period has to be even smaller, which greatly increases the processing difficulty and cost of the device. This is also the main reason why there is little research on multi-band absorbers in the visible band currently. Summary of the Invention

[0005] Aiming at the deficiencies of the existing technology, the present invention aims to provide a multispectral imaging food detection device based on a narrow-band absorber in the visible light band. The narrow-band absorber designed in the present invention has continuously adjustable absorption peaks, so it can be applied to multispectral imaging of different fruits, vegetables, and root vegetables.

[0006] To solve the above technical problems, the technical solution adopted by the present invention is: a multispectral imaging food detection device based on a narrow-band absorber in the visible light band, including a laser-driven broadband light source, a semi-transparent and semi-reflective mirror, a detection platform, a polarizer, a plane mirror, a multispectral imaging sensor, and an image processing system;

[0007] The laser-driven broadband light source emits continuous light in the band of 300nm - 1200nm, which is irradiated on the plant-based food placed on the detection platform through the semi-transparent and semi-reflective lens. The reflected light is collected by the objective lens and then passes through the semi-transparent and semi-reflective lens and the polarizer in sequence to obtain TM-polarized polarized light. Finally, the TM-polarized polarized light is reflected by the plane mirror and enters the multispectral imaging sensor. The pixels of the multispectral imaging sensor are multispectral narrow-band absorbers, and the image transmitted to the image processing system through the signal is an image formed by the superposition of the characteristic wavelengths of the detected object.

[0008] The structure of the multispectral narrow-band absorber adopts a sandwich waveguide structure formed by metal-dielectric-metal. The bottom substrate is silver, the middle layer uses silicon dioxide, and the surface uses a silver grating. The grating is periodically arranged along the x direction. Each microstructure unit period includes three silver gratings, and the spacing between the silver gratings is different.

[0009] Furthermore, the width and height of each silver grating are 80 nm and 130 nm respectively.

[0010] Furthermore, the period of the minimum unit of the microstructure is 500 nm.

[0011] The advantages and positive effects of the present invention are as follows: The present invention designs a multispectral imaging system based on a visible-light narrowband absorber, arranges the designed absorbers into a two-dimensional array as the imaging unit of the detector to directly form a visual image. The microstructure of the absorber adopts a sandwich-structured waveguide formed by metal-dielectric-metal, and the top and bottom metal layers can be directly used as electrodes and integrated into the electro-optical system, miniaturizing the multispectral imaging system. The absorption peak of the narrowband absorber designed in the present invention is continuously adjustable, so it can be applied to the multispectral imaging of different fruits, vegetables and root vegetables. Description of the Drawings

[0012] Figure 1 It is the schematic diagram of the multispectral imaging of the present invention.

[0013] Figure 2 It is the microstructure diagram of the narrowband absorber of the present invention.

[0014] Figure 3 It is the absorption rate chromaticity diagram of the absorption peak varying with the dielectric thickness (d1 = 30 nm, d2 = 80 nm).

[0015] Figure 4 It is the spectral diagram of the absorption rate and reflectance in the visible light band (t die = 428 nm).

[0016] Figure 5 It is the cross-sectional diagram of the magnetic field distribution of six absorption peaks in the microstructure. Detailed Embodiment

[0017] To make the objectives, features and advantages of the present invention more obvious and understandable, the following provides a detailed description of the specific embodiments of the present invention.

[0018] As Figure 1 shown, the multispectral imaging food detection device based on the visible-light band narrowband absorber includes a laser-driven broadband light source, a semi-transmissive and semi-reflective mirror, a detection platform, a polarizer, a plane mirror, a multispectral imaging sensor and an image processing system.

[0019] The laser-driven broadband light source emits continuous light in the wavelength range of 300 nm to 1200 nm. The light irradiates the plant-based food placed on the detection platform through a semi-reflective and semi-transmissive lens. The reflected light is collected by an objective lens and then passes through the semi-reflective and semi-transmissive lens and a polarizer in sequence to obtain TM-polarized polarized light. Finally, the TM-polarized polarized light is reflected by a plane mirror and enters a multi-spectral imaging sensor. The pixels of the multi-spectral imaging sensor are multi-spectral narrow-band absorbers. The image transmitted to the image processing system through the signal is an image formed by the superposition of the characteristic wavelengths of the detected object. The peak wavelength of the visible light narrow-band absorber can be matched with the characteristic wavelength of the detected food by adjusting the structural parameters.

[0020] The structural diagram of the narrow-band absorber in the visible light band is as Figure 2 shown. This structure adopts a sandwich waveguide structure formed by metal-dielectric-metal. The bottom substrate uses the good conductor metal silver (Ag) with relatively rich mineral resources and moderate price, and its thickness is 100 nm. The middle layer uses silicon dioxide (SiO2) which is rich in resources and has high quantum efficiency in the visible light band. The surface uses a silver grating, and the grating is periodically arranged in the x direction, and its width and height are 80 nm and 130 nm respectively. The period of the minimum unit of the microstructure is 500 nm. In order to achieve multi-spectral narrow-band absorption, the spacings of the three silver gratings within the unit period are different. By adjusting the period of the microstructure unit, the grating spacing and the thickness of the dielectric layer, the absorption peak can be adjusted so that the multi-spectral imaging technology based on the visible light narrow-band absorber is applicable to detecting different types of plant-based foods.

[0021] Figure 3 shows the absorption rate chromaticity diagram of the narrow-band absorber in the visible light band when d1 and d2 are equal to 30 nm and 80 nm respectively, and the absorption peak changes with the dielectric thickness. The chromaticity range is 0 to 1, and 0 and 1 represent 0% absorption and 100% absorption respectively. It can be seen that this narrow-band absorber can achieve 6 narrow-band absorption peaks, and the first 4 absorption peaks show a red shift of the absorption peak with the increase of the electrolyte thickness in the visible light range.

[0022] Figure 4 shows the spectral diagram of the transmittance, absorption rate and reflectance in the visible light band with the change of wavelength when the dielectric thickness is 428 nm. The red line is the absorption rate, the blue line is the reflectance, and the green line is the transmittance. Since the bottom layer of the microstructure is a silver film and its thickness is greater than the skin depth of visible light, the transmittance is 0. Observing the red line, it can be seen that the absorption rates of the 6 absorption peaks (532 nm, 556 nm, 652 nm, 702 nm, 750 nm, 770 nm) all exceed 60%, and the absorption rates of two of the peaks are close to 90%. And the bandwidth of the absorption peak is 2 nm to 8 nm. Compared with the absorption peaks of the currently studied infrared absorbers which are 2 to 4, the absorber designed in the present invention has more absorption peaks and realizes narrow-band absorption in the visible light range.

[0023] AsFigure 5 The cross-sectional view of the magnetic field distribution in the microstructure shows six absorption peaks. Figure 5 (a) and Figure 5 (c) show that strong absorption of the electromagnetic field is generated between the metal slits. This is because a resonance cavity is formed at the metal-air-metal (MAM) interface, causing resonance of light with wavelengths of 532 nm and 652 nm. The absorption intensity of the MAM cavity is related to the width d1 of the metal slit. When the slit width exceeds 30 nm, this absorption peak will disappear. Figure 5 (b), (d), and (f) show that the magnetic field forms second-order, first-order, and zero-order FT resonance modes in the middle and upper layers of the metal-dielectric-metal structure, and the corresponding three absorption peaks are 556 nm, 702 nm, and 770 nm. Figure 5 (e) is a typical surface plasmon mode occurring at the metal-dielectric interface. Its absorption peak is only related to the grating period P. When the grating period becomes longer, its absorption peak will undergo a red shift.

[0024] Compared with the hyperspectral sensor, where the acquisition and processing of hundreds of consecutive narrow-band images limit the speed of food safety detection, the multispectral imaging technology based on narrow-band absorbers in the visible light band can achieve rapid detection, and the detection speed can be increased by at least 5 times. Compared with the narrow-band absorbers in the multispectral imaging system, which are mostly applied in the microwave, terahertz, and infrared bands, and the number of absorption peaks is mostly 2 - 4. The narrow-band absorber in the present invention can achieve multispectral narrow-band absorption in the visible band, and the number of absorption peaks is six, and the narrowest bandwidth is 2 nm, achieving high spectral resolution and can be applied in the field of food safety detection.

[0025] The above has described the embodiments of the present invention in detail. However, the present invention is not limited to the above embodiments. Within the scope of knowledge possessed by those of ordinary skill in the art, various changes can be made without departing from the gist of the present invention.

Claims

1. A multispectral imaging food detection device based on a narrowband absorber in the visible light band, characterized in that: It includes a laser-driven broadband light source, a semi-transparent and semi-reflective mirror, a detection platform, a polarizer, a plane mirror, a multispectral imaging sensor, and an image processing system; The laser-driven broadband light source emits continuous light in the wavelength band of 300 nm to 1200 nm. The light irradiates the plant-based food placed on the detection platform through the semi-transparent and semi-reflective lens. The reflected light is collected by the objective lens and then passes through the semi-transparent and semi-reflective lens and the polarizer in sequence to obtain TM-polarized polarized light. Finally, the TM-polarized polarized light is reflected by the plane mirror and enters the multispectral imaging sensor. The pixel of the multispectral imaging sensor is a multispectral narrowband absorber. The image transmitted to the image processing system through the signal is an image formed by superimposing the characteristic wavelengths of the detected object; The structure of the multispectral narrowband absorber adopts a sandwich waveguide structure formed by metal-dielectric-metal. The bottom substrate is silver, the middle layer uses silicon dioxide, and the surface uses a silver grating. The grating is periodically arranged in the x direction. Each microstructure unit period includes three silver gratings, and the distances between the silver gratings are different; The width and height of each silver grating are 80 nm and 130 nm respectively; The minimum unit period of the microstructure is 500 nm; In the visible light band, the narrowband absorber has d1 and d2 equal to 30 nm and 80 nm respectively. The d1 is the distance between the first silver grating and the second silver grating among the three silver gratings arranged in sequence within the microstructure unit period, and d2 is the distance between the second silver grating and the third silver grating among the three silver gratings arranged in sequence within the microstructure unit period.

Images