A probe and a method and device for detecting hazardous objects based on deep learning
Through the technology combined with tri-color fluorescence probe and deep learning method, the sensitivity and complexity of fluorescence sensors when detecting Hg2+ and S2- are solved, and high sensitivity, visual detection and rapid identification of heavy metals are achieved, suitable for a variety of applications Water detection in the environment.
Patent Information
- Application Number
- CN202210684213.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-06-16
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2042-06-16
AI Technical Summary
When existing fluorescent sensors detect heavy metals Hg2+ and sulfide S2-, single-emitting probes are greatly affected by probe concentration, instrument noise and sample environment, and are complex in manufacturing. The color changes of the dual-emitting probes are small, making it difficult to achieve significant color changes.
A tricolor fluorescent probe consisting of blue, green and red emitted zinc telluride quantum dots is used to achieve dual fluorescence quenching of Hg2+ and S2- through electron transfer effect, and fluorescence image analysis is performed in combination with deep learning methods to identify RGB signal to determine the concentration.
Realize visual detection of Hg2+ and S2-, with excellent selectivity and ultra-high sensitivity, simplifies the probe preparation process, and can quickly and accurately identify targets in complex environments. It is suitable for agricultural water, domestic water and seafood. Test.
Smart Images

Figure CN115078322B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of photochemical technology, and specifically relates to a probe and a method and device for detecting hazardous substances based on deep learning. Background Art
[0002] So far, heavy metal pollution has become an increasingly serious environmental pollution problem. 2+ Hg is a widely distributed pollutant with unique toxicological properties. It is a corrosive carcinogen with high cytotoxicity. Due to the influence of volcanic activity, mining and industrial production, Hg in the environment 2+ Pollution is increasing. However, Hg 2+ It cannot be naturally degraded, so Hg 2+ It is bioaccumulative, which can cause serious biological and environmental problems. So far, Hg has been detected in a wide range of soil, water, and food. 2+ Hg has even been detected in some animal blood and plant biological cells. 2+ In addition, sulfide ions (S 2- ) is also a hazardous ion widely present in natural water and wastewater systems. It is mainly produced by various production processes such as sulfur and sulfuric acid production, dye and cosmetic production, and metal precipitation in wastewater treatment. At the same time, sulfide is also produced by the microbial reduction of sulfate by anaerobic bacteria. Sulfide is a toxic pollutant that is harmful to human health and the environment. Therefore, S 2- The concentration of Hg 2+ and S 2- Real-time monitoring and detection in each cycle link is a scientific problem that needs to be solved urgently.
[0003] Currently, Hg 2+ and S 2- Great progress has been made in the quantitative research and analysis of Hg 2+ and S 2- Commonly used analytical methods include: atomic absorption, fluorescence spectroscopy, inductively coupled plasma mass spectrometry, surface enhanced Raman spectroscopy (SERS), chemiluminescence, localized surface plasmon resonance (LSPR) and electrochemical methods. Among the above methods, fluorescence analysis is considered to be the most promising technology due to its high sensitivity and high selectivity. In addition, the fluorescence sensing method used in fluorescence analysis is also a simple, fast and effective analytical method because such methods do not require complex sample preparation before analysis and detection. Fluorescence sensing can be constructed as "on", "off" and "ratio" response types. At present, the existing technology has developed several methods for the determination of Hg 2+ and S 2-The construction of these fluorescent sensor systems mainly relies on fluorescent nanomaterials, including metal clusters (M-NCs), quantum dots (QDs), luminescent metal-organic frameworks (M-MOFs), etc.
[0004] However, among many fluorescent sensors, single-emission fluorescent probes with independent fluorescence signals are greatly affected by probe concentration, instrument noise and sample environment, and are not easy to show obvious color changes. In contrast, the construction of dual-color fluorescent probes based on ratiometric fluorescence sensing can significantly enhance color contrast and improve detection sensitivity. For example, Liu Wei et al. constructed a dual-emission fluorescent probe C-dot@SiO2@Q-dots with SiO2 as the connecting layer to detect Cu 2+ Li Suying et al. constructed N,Fe-CDs / BSA-Ag-NCs ratiometric fluorescence complex with peroxidase activity for the detection of uric acid and hydrogen peroxide. However, the dual-emission probe not only has a complex manufacturing process, but also has a small color change. In order to solve this problem, it is imperative to develop a multicolor fluorescence sensing system. Summary of the invention
[0005] The present invention aims to solve the technical problems that the single-emission fluorescent probe in the prior art is greatly affected by the probe concentration, instrument noise and sample environment and is not easy to show obvious color change, and the dual-emission probe not only has a complex manufacturing process but also has a small color change. A probe is provided, and the probe is prepared by the following method:
[0006] Dissolve citric acid and cysteamine hydrochloride in ultrapure water to obtain solution A. After sterilization, solution A is heated at 150-180° C. for 3-6 hours, cooled, and purified to obtain blue emitting carbon quantum dots.
[0007] The cadmium salt, zinc salt and stabilizer are dissolved in ultrapure water to obtain solution B. The pH of solution B is adjusted to 8-9 to obtain solution C. Solution C is degassed by bubbling in an ice bath in an inert gas environment for 25-45 minutes, and sodium telluride precursor solution is added under vigorous stirring, and refluxed at 92-96° C. for 22-28 hours to obtain green-emitting cadmium zinc telluride quantum dots.
[0008] The cadmium salt, zinc salt and stabilizer are dissolved in ultrapure water to obtain solution B. The pH of solution B is adjusted to 8-9 to obtain solution C. Solution C is degassed by bubbling in an ice bath in an inert gas environment for 25-45 minutes, and sodium telluride precursor solution is added under vigorous stirring, and refluxed at 92-96° C. for 160-170 hours to obtain red-emitting cadmium zinc telluride quantum dots.
[0009] The probe is obtained by mixing the above-mentioned blue emitting carbon quantum dots, green emitting cadmium zinc telluride quantum dots and red emitting cadmium zinc telluride quantum dots.
[0010] Furthermore, the mass concentration of citric acid in the solution A is 40-45 g / L. The mass concentration of cysteamine hydrochloride in the solution A is 45-50 g / L.
[0011] Furthermore, the cadmium salt is cadmium chloride, the zinc salt is zinc chloride, and the stabilizer is trimercaptopropionic acid.
[0012] Furthermore, the mass concentration of the cadmium salt in the solution B is 450-460 mg / L, the mass concentration of the zinc salt in the solution B is 400-410 mg / L, and the mass concentration of the stabilizer in the solution B is 1-2 g / L.
[0013] Furthermore, the method for adjusting the pH of solution B to 8-9 is: adding 1 mol / L sodium hydroxide solution dropwise to solution B until the pH of solution B is adjusted to 8-9.
[0014] Furthermore, when the blue-emitting carbon quantum dots, green-emitting cadmium zinc telluride quantum dots, and red-emitting cadmium zinc telluride quantum dots are mixed, the volume ratio of the blue-emitting carbon quantum dots, the green-emitting cadmium telluride quantum dots, and the red-emitting cadmium telluride quantum dots is blue-emitting carbon quantum dots: green-emitting cadmium telluride quantum dots: red-emitting cadmium telluride quantum dots = 14-18:100:12-16.
[0015] In addition, the present invention also provides a hazard detection method based on deep learning, comprising:
[0016] S1. Obtain fluorescence image information and divide the fluorescence images into a training set and a test set.
[0017] S2. In the training set, label all test tubes and annotate the training set in VOC2007 format.
[0018] S3. Use the YOLO v3 algorithm to make predictions and extract the RGB signal values of the test tube positions and test tube colors in the training set annotated in the VOC2007 format.
[0019] S4. Segment and extract features of the predicted fluorescence image to complete the training of the YOLO v3 algorithm.
[0020] S5. Use the YOLO v3 algorithm to extract the RGB signal values of the test tube positions and test tube colors in the test set and generate a linear relationship.
[0021] S6. Obtain the concentrations of Hg2+ and S2- based on the linear relationship.
[0022] At the same time, the present invention also provides a hazard detection device based on deep learning, including an intelligent terminal, a dark box and a portable shooting device.
[0023] The smart terminal is used to receive data, store data and analyze data. The smart terminal uses the above-mentioned hazard detection method based on deep learning to analyze the received data.
[0024] The dark box comprises: an opaque box body, a placement slot for installing a portable camera device is provided on one side of the box body. A multi-test tube holder is detachably installed on the other side of the placement slot of the box body. The test tubes on the multi-test tube holder are filled with the above-mentioned probes. An ultraviolet lamp is provided above the interior of the box body. Through holes are provided on the box body at positions corresponding to the placement slot and the multi-test tube holder, which allow the portable camera device to obtain probe images.
[0025] The portable shooting device is installed in the placement slot of the dark box, and obtains the fluorescent image information of the probes in the multi-test tube support through the through hole, and sends it to the intelligent terminal.
[0026] The present invention has at least one of the following advantages:
[0027] 1. The present invention designs and constructs a three-color sensing probe, which has the following main advantages: (1) The preparation of the three-color probe is simple and does not require other cumbersome steps. (2) Rich and extensive color changes can be observed with the naked eye. (3) The three-color probe can detect Hg 2+ and S 2- It has excellent selectivity and ultra-high sensitivity.
[0028] 2. The probe of the present invention is a mixture of B-QDs, G-QDs and R-QDs, and has rich color changes. 2+ and S 2- The increase in the content will lead to the double quenching of the fluorescence intensity of R-QDs and G-QDs based on the electron transfer effect, but will not affect the fluorescence intensity of B-QDs, thus achieving the 2+ and S 2- Visual inspection.
[0029] 3. The present invention can capture fluorescent colorimetric images through a camera device, identify and process photos, and use software to cut and read RGB values to test Hg in real samples 2+ and S 2- The deep learning function based on the YOLO v3 algorithm and the end-to-end target detection method make the hazard detection device based on deep learning of the present invention have the advantages of fast recognition speed, high recognition accuracy, and strong model anti-interference ability. It can have the ability to identify targets in complex environments. The sensing system realizes the detection of Hg in agricultural water, domestic water and seafood. 2+ and S 2- Highly sensitive detection.
[0030] 4. The portable optical device (dark box) provided by the present invention combines deep learning technology with a smartphone to capture RGB values, which can be used flexibly, conveniently and accurately for Hg 2+ and S 2- Visual detection of Hg content in the monitoring target 2+ and S 2- Regular monitoring of content. BRIEF DESCRIPTION OF THE DRAWINGS
[0031] Figure 1 Shown is the Hg 2+ Curve change trend and fluorescence color change diagram during detection.
[0032] Figure 2 The present invention is shown in S 2- Curve change trend and fluorescence color change diagram during detection. DETAILED DESCRIPTION
[0033] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.
[0034] Example 1
[0035] A probe is prepared by the following method:
[0036] Dissolve citric acid and cysteamine hydrochloride in ultrapure water to obtain solution A. After sterilization, solution A is heated at 160°C for 4 hours, cooled, and purified to obtain blue emitting carbon quantum dots. The mass concentration of citric acid in solution A is 42 g / L. The mass concentration of dry Guangan hydrochloride in solution A is 46 g / L.
[0037] Cadmium chloride, zinc chloride and trimercaptopropionic acid were dissolved in ultrapure water to obtain solution B. The mass concentration of cadmium chloride in solution B was 456 mg / L, the mass concentration of zinc chloride in solution B was 408 mg / L, and the mass concentration of trimercaptopropionic acid in solution B was 1.3 g / L. 1 mol / L sodium hydroxide solution was added dropwise to solution B until the pH of solution B was adjusted to 8.2 to obtain solution C. Solution C was degassed by bubbling in an ice bath in an inert gas environment for 30 min, and sodium telluride precursor solution was added under vigorous stirring, and refluxed at 95° C. for 24 h to obtain green-emitting cadmium zinc telluride quantum dots.
[0038] Cadmium chloride, zinc chloride and trimercaptopropionic acid were dissolved in ultrapure water to obtain solution B. The mass concentration of cadmium chloride in solution B was 456 mg / L, the mass concentration of zinc chloride in solution B was 408 mg / L, and the mass concentration of trimercaptopropionic acid in solution B was 1.3 g / L. 1 mol / L sodium hydroxide solution was added dropwise to solution B until the pH of solution B was adjusted to 8.2 to obtain solution C. Solution C was degassed by bubbling in an ice bath in an inert gas environment for 30 min, and sodium telluride precursor solution was added under vigorous stirring, and refluxed at 95° C. for 168 h to obtain red-emitting cadmium zinc telluride quantum dots.
[0039] The volume ratio of blue emission carbon quantum dots, green emission cadmium telluride quantum dots and red emission cadmium telluride quantum dots is: blue emission carbon quantum dots: green emission cadmium telluride quantum dots: red emission cadmium telluride quantum dots
[0040] The probe is obtained by mixing the blue emitting carbon quantum dots, the green emitting cadmium zinc telluride quantum dots, and the red emitting cadmium zinc telluride quantum dots in a ratio of 1:16:100:15.
[0041] Example 2
[0042] A probe is prepared by the following method:
[0043] Dissolve citric acid and cysteamine hydrochloride in ultrapure water to obtain solution A. After sterilization, solution A is heated at 180°C for 6 hours, cooled, and purified to obtain blue emitting carbon quantum dots. The mass concentration of citric acid in solution A is 45 g / L. The mass concentration of dry Guangan hydrochloride in solution A is 50 g / L.
[0044] Cadmium chloride, zinc chloride and trimercaptopropionic acid were dissolved in ultrapure water to obtain solution B. The mass concentration of cadmium chloride in solution B was 460 mg / L, the mass concentration of zinc chloride in solution B was 410 mg / L, and the mass concentration of trimercaptopropionic acid in solution B was 2 g / L. 1 mol / L sodium hydroxide solution was added dropwise to solution B until the pH of solution B was adjusted to 9 to obtain solution C. Solution C was degassed by bubbling in an ice bath in an inert gas environment for 45 min, and sodium telluride precursor solution was added under vigorous stirring, and refluxed at 96° C. for 28 h to obtain green-emitting zinc telluride cadmium quantum dots.
[0045] Cadmium chloride, zinc chloride and trimercaptopropionic acid were dissolved in ultrapure water to obtain solution B. The mass concentration of cadmium chloride in solution B was 460 mg / L, the mass concentration of zinc chloride in solution B was 410 mg / L, and the mass concentration of trimercaptopropionic acid in solution B was 2 g / L. 1 mol / L sodium hydroxide solution was added dropwise to solution B until the pH of solution B was adjusted to 9 to obtain solution C. Solution C was degassed by bubbling in an ice bath in an inert gas environment for 45 min, and sodium telluride precursor solution was added under vigorous stirring, and refluxed at 96° C. for 170 h to obtain red-emitting cadmium zinc telluride quantum dots.
[0046] The volume ratio of blue-emitting carbon quantum dots, green-emitting cadmium telluride quantum dots and red-emitting cadmium telluride quantum dots is blue-emitting carbon quantum dots: green-emitting cadmium telluride quantum dots: red-emitting cadmium telluride quantum dots = 18:100:16. The probe is obtained by mixing the above-mentioned blue-emitting carbon quantum dots, green-emitting cadmium zinc telluride quantum dots and red-emitting cadmium zinc telluride quantum dots.
[0047] Example 3
[0048] A probe is prepared by the following method:
[0049] Dissolve citric acid and cysteamine hydrochloride in ultrapure water to obtain solution A. After sterilization, solution A is heated at 150°C for 3 hours, cooled, and purified to obtain blue emitting carbon quantum dots. The mass concentration of citric acid in solution A is 40 g / L. The mass concentration of dry Guangan hydrochloride in solution A is 45 g / L.
[0050] Cadmium chloride, zinc chloride and trimercaptopropionic acid are dissolved in ultrapure water to obtain solution B. The mass concentration of cadmium chloride in solution B is 450 mg / L, the mass concentration of zinc chloride in solution B is 400-410 mg / L, and the mass concentration of trimercaptopropionic acid in solution B is 1 g / L. 1 mol / L sodium hydroxide solution is added dropwise to solution B until the pH of solution B is adjusted to 8 to obtain solution C. Solution C is degassed by bubbling in an ice bath in an inert gas environment for 25 min, and sodium telluride precursor solution is added under vigorous stirring, and refluxed at 92° C. for 22 h to obtain green-emitting cadmium zinc telluride quantum dots.
[0051] Cadmium chloride, zinc chloride and trimercaptopropionic acid are dissolved in ultrapure water to obtain solution B. The mass concentration of cadmium chloride in solution B is 450 mg / L, the mass concentration of zinc chloride in solution B is 400 mg / L, and the mass concentration of trimercaptopropionic acid in solution B is 1 g / L. 1 mol / L sodium hydroxide solution is added dropwise to solution B until the pH of solution B is adjusted to 8 to obtain solution C. Solution C is degassed by bubbling in an ice bath in an inert gas environment for 25 min, and sodium telluride precursor solution is added under vigorous stirring, and refluxed at 92° C. for 160 h to obtain red-emitting cadmium zinc telluride quantum dots.
[0052] The volume ratio of blue-emitting carbon quantum dots, green-emitting cadmium telluride quantum dots and red-emitting cadmium telluride quantum dots is blue-emitting carbon quantum dots: green-emitting cadmium telluride quantum dots: red-emitting cadmium telluride quantum dots = 14:100:12. The probe is obtained by mixing the above-mentioned blue-emitting carbon quantum dots, green-emitting cadmium zinc telluride quantum dots and red-emitting cadmium zinc telluride quantum dots.
[0053] The probe of the present invention is composed of blue-emitting carbon quantum dots (B-QDs), green-emitting cadmium zinc telluride quantum dots (G-QDs) and red-emitting cadmium zinc telluride quantum dots (R-QDs). 2+ and S 2- The responses of G-QDs and R-QDs are based on the dynamic quenching principle and the static quenching principle, respectively. In this three-color sensing system, the fluorescence intensity of G-QDs and R-QDs is simultaneously 2+ and S 2- quenching, while B-QDs 2+ and S 2- Insensitive, used as an internal reference. 2+ and S 2- When the probe is exposed to light, the fluorescence color of the probe changes from green to red and then to blue.
[0054] The probe of the present invention is a mixture of B-QDs, G-QDs and R-QDs, and has rich color changes. The preparation method of the probe of the present invention is not complicated, and does not involve complex preparation technology and cumbersome preparation process. 2+ and S 2- Excellent selectivity for Hg 2+ and S 2- The content of Hg is highly sensitive. 2+ and S 2- The increase in the content will lead to the double quenching of the fluorescence intensity of R-QDs and G-QDs based on the electron transfer effect, but will not affect the fluorescence intensity of B-QDs. Therefore, the probe of the present invention can achieve the detection of Hg 2+ and S 2- Visual inspection.
[0055] Taking a certain test as an example, Hg 2+ The detection range is 0.004-97.15μM, the lod is 0.006μM, the curve change trend and fluorescence color change are as follows Figure 1 As shown. 2- The detection range is 5-480μM, the lod is 3.576μM, the curve change trend and fluorescence color change are as follows Figure 2 As shown. Figure 1 and Figure 2 It can be seen that the probe of the present invention can accurately identify Hg 2+ and S 2- , and with Hg 2+ and S 2- The color of the probe changes significantly with the change of Hg content. 2+ and S 2- Excellent selectivity for Hg 2+ and S 2- The content has ultra-high sensitivity.
[0056] Example 4
[0057] A hazard detection method based on deep learning, comprising:
[0058] S1. Obtain fluorescence image information and divide the fluorescence images into a training set and a test set.
[0059] S2. In the training set, label all test tubes and annotate the training set in VOC2007 format.
[0060] S3. Use the YOLO v3 algorithm to make predictions and extract the RGB signal values of the test tube positions and test tube colors in the training set annotated in the VOC2007 format.
[0061] S4. Segment and extract features of the predicted fluorescence image to complete the training of the YOLO v3 algorithm.
[0062] S5. Use the YOLO v3 algorithm to extract the RGB signal values of the test tube positions and test tube colors in the test set and generate a linear relationship.
[0063] S6. Obtain Hg based on the linear relationship 2+ and S 2- concentration.
[0064] Example 5
[0065] A hazard detection device based on deep learning comprises an intelligent terminal, a dark box and a portable shooting device.
[0066] The smart terminal is used to receive data, store data and analyze data. The smart terminal uses the above-mentioned hazard detection method based on deep learning to analyze the received data.
[0067] The dark box comprises: an opaque box body, a placement slot for installing a portable camera device is provided on one side of the box body. A multi-test tube holder is detachably installed on the other side of the placement slot of the box body. The test tubes on the multi-test tube holder are filled with the above-mentioned probes. An ultraviolet lamp is provided above the interior of the box body. Through holes are provided on the box body at positions corresponding to the placement slot and the multi-test tube holder, which allow the portable camera device to obtain probe images.
[0068] The portable shooting device is installed in the placement slot of the dark box, and obtains the fluorescence image information of the probe in the multi-test tube rack through the through hole, and sends it to the smart terminal. The portable shooting device can be a smart phone with a camera and camera software as needed.
[0069] Based on the probe of the present invention, Hg 2+ and S 2- Excellent selectivity for Hg 2+ and S 2- The content of Hg in agricultural water, domestic water and seafood has ultra-high sensitivity. Combined with YOLO v3 model analysis, this hazard detection device based on deep learning can achieve 2+ and S 2- Highly sensitive and accurate detection, so that Hg in agricultural water, domestic water and seafood can be detected by regular testing. 2+ and S 2- Monitor and analyze the content.
[0070] In addition, YOLO v3 is an end-to-end target detection method, which enables the hazard detection device based on deep learning of the present invention to have the advantages of fast recognition speed, high recognition accuracy, strong model anti-interference ability, etc., and can have the ability to identify targets in complex environments.
[0071] The hazard detection device based on deep learning of the present invention can quickly identify Hg in agricultural water, domestic water and seafood. 2+ and S 2- content, so that it is convenient for the processing unit to analyze the Hg in agricultural water, domestic water and seafood 2+ and S 2- Carry out targeted treatment to reduce Hg in agricultural and domestic water 2+ and S 2- Entering the human body through contaminated crops and Hg in seafood 2+ and S 2-The possibility of entering the human body after consumption. Achieve the technical purpose of improving the safety of food, especially crops, fishery products, etc.
[0072] It should be noted and understood that various modifications and improvements can be made to the invention described in detail above without departing from the spirit and scope of the claims of the present invention. Therefore, the scope of the claimed technical solution is not limited by any specific exemplary teaching given.
Claims
1. A probe, characterized in that: The probe is used to detect Hg 2+ and S 2- , and prepared by the following method: Dissolving citric acid and cysteamine hydrochloride in ultrapure water to obtain solution A; sterilizing solution A, heating it at 150-180° C. for 3-6 hours, cooling it, and purifying it to obtain blue emitting carbon quantum dots; Dissolve cadmium salt, zinc salt and stabilizer in ultrapure water to obtain solution B; adjust the pH of solution B to 8-9 to obtain solution C; degas solution C in an inert gas environment by bubbling in an ice bath for 25-45 minutes, add sodium telluride precursor solution under vigorous stirring, and reflux at 92-96° C. for 22-28 hours to obtain green-emitting cadmium zinc telluride quantum dots; Dissolve cadmium salt, zinc salt and stabilizer in ultrapure water to obtain solution B; adjust the pH of solution B to 8-9 to obtain solution C; degas solution C in an ice bath in an inert gas environment for 25-45 minutes, add sodium telluride precursor solution under vigorous stirring, and reflux at 92-96° C. for 160-170 hours to obtain red-emitting cadmium zinc telluride quantum dots; The probe is obtained by mixing the above-mentioned blue emitting carbon quantum dots, green emitting cadmium zinc telluride quantum dots, and red emitting cadmium zinc telluride quantum dots; The mass concentration of citric acid in the solution A is 40-45 g / L; the mass concentration of cysteamine hydrochloride in the solution A is 45-50 g / L; The cadmium salt is cadmium chloride, the zinc salt is zinc chloride, and the stabilizer is trimercaptopropionic acid; The mass concentration of the cadmium salt in the solution B is 450-460 mg / L, the mass concentration of the zinc salt in the solution B is 400-410 mg / L, and the mass concentration of the stabilizer in the solution B is 1-2 g / L; When the blue-emitting carbon quantum dots, green-emitting cadmium zinc telluride quantum dots and red-emitting cadmium zinc telluride quantum dots are mixed, the volume ratio of the blue-emitting carbon quantum dots, green-emitting cadmium telluride quantum dots and red-emitting cadmium telluride quantum dots is: blue-emitting carbon quantum dots: green-emitting cadmium telluride quantum dots: red-emitting cadmium telluride quantum dots = 14-18:100:12-16.
2. The probe according to claim 1, characterized in that: The method for adjusting the pH of solution B to 8-9 is: adding 1 mol / L sodium hydroxide solution dropwise to solution B until the pH of solution B is adjusted to 8-9.
3. A pollutant detection device based on deep learning, characterized in that: It includes an intelligent terminal, a dark box and a portable shooting device; The intelligent terminal is used to receive data, store data and analyze data; The intelligent terminal uses a pollutant detection method based on deep learning to perform data analysis on the received data; The pollutant detection method based on deep learning includes: S1. Obtaining fluorescence image information and dividing the fluorescence images into a training set and a test set; S2. In the training set, label all test tubes and annotate the training set in VOC2007 format; S3. Use the YOLO v3 algorithm to make predictions and extract the RGB signal values of the test tube positions and test tube colors in the training set annotated in the VOC2007 format; S4. Segment and extract features of the predicted fluorescence image to complete the training of the YOLO v3 algorithm; S5. Use the YOLO v3 algorithm to extract the RGB signal values of the test tube positions and test tube colors in the test set and generate a linear relationship; S6. Obtain Hg based on the linear relationship 2+ and S 2- concentration; The dark box comprises: an opaque box, one side of which is provided with a placement slot for installing a portable camera; a multi-test tube holder is detachably installed on the other side of the placement slot; the test tubes on the multi-test tube holder are filled with the probes according to any one of claims 1-2; an ultraviolet lamp is arranged above the interior of the box; and through holes are arranged on the box at positions corresponding to the placement slot and the multi-test tube holder, which allow the portable camera to obtain probe images; The portable shooting device is installed in the placement slot of the dark box, and obtains the fluorescent image information of the probes in the multi-test tube support through the through hole, and sends it to the intelligent terminal.
