Microalgae feature fusion analysis method and system based on deep learning

Abstract

The invention relates to a microalgae feature fusion analysis method and system based on deep learning, and belongs to the technical field of microalgae feature recognition and analys.The microalgae collection time node is obtained according to the microalgae growth collection threshold range, and the real-time pollutant data of the current microalgae collection time node are obtained; and finally, pollution analysis is performed according to the real-time pollutant data of the current microalgae harvesting time node, a related regulation and control strategy is generated based on a pollution analysis result, and regulation and control are performed according to the related regulation and control strategy. According to the method, the deep learning network and the generative adversarial network are fused, the cell division laws of the microalgae in different environments and the image features within the preset time can be fused to analyze the microalgae growth data in the target area, and the problem of the pain point of the image technology for microalgae growth and reproduction recognition is solved; the growth stage of the microalgae is accurately grasped, and a theoretical support is provided for identifying, regulating and removing microalgae pollutants in a watershed water body.

Description

Technical Field

[0001] The present invention relates to the technical field of microalgae identification and analysis, and in particular to a microalgae feature fusion analysis method and system based on deep learning. Background Art

[0002] Microalgae can convert solar energy into renewable energy materials such as hydrogen, hydrocarbons, alcohols, and oils and store them in cells. As raw materials for biodiesel, alcohol-based fuels, biohydrogen production technology, microalgae hydrocarbon production, biofertilizers, and microalgae plastics, they can meet the needs of industrial and agricultural applications to a certain extent. However, due to the small size of microalgae cells, the density close to that of water, and the negative charge on the surface, it is difficult for microalgae to settle naturally and separate from water, making the cost of microalgae harvesting account for 20% to 30% of the total production cost, which seriously limits its large-scale and industrial production. Traditional microalgae harvesting technologies include physical methods (such as centrifugation, membrane filtration, and flotation), chemical methods (such as flocculation), and electromagnetic methods (such as electroflocculation and magnetic flocculation). The application of these technologies generally requires a lot of energy consumption and high operating costs to drive, and is easy to cause damage to microalgae cells. Therefore, it is necessary to use image recognition technology, spectral technology, etc. for identification. However, due to the lack of microalgae image data, the limitations of image technology, and changes in the culture environment, it is impossible to truly predict the reproduction of microalgae, resulting in the failure of microalgae to meet the predetermined requirements when harvested. Summary of the invention

[0003] The present invention overcomes the deficiencies of the prior art and provides a method and system for fusion analysis of microalgae features based on deep learning.

[0004] To achieve the above object, the technical solution adopted by the present invention is: The first aspect of the present invention provides a method for fusion analysis of microalgae features based on deep learning, comprising the following steps: Acquire image feature data of microalgae culture solution in the target area, build a microalgae growth prediction model based on a deep neural network, and optimize the microalgae growth prediction model through a generative adversarial network; Predicting the image feature data of the microalgae culture solution in the target area by using the microalgae growth prediction model to obtain microalgae growth data within a preset time; Setting a microalgae growth and harvesting threshold range, obtaining a microalgae harvesting time node according to the microalgae growth data within the preset time and the microalgae growth and harvesting threshold range, and obtaining real-time pollutant data at the current microalgae harvesting time node; Pollution analysis is performed according to the real-time pollutant data at the current microalgae harvesting time node, and relevant control strategies are generated based on the pollution analysis results, and control is performed according to the relevant control strategies.

[0005] Furthermore, in the microalgae feature fusion analysis method based on deep learning, the image feature data of the microalgae culture solution in the target area is obtained, and a microalgae growth prediction model is constructed based on a deep neural network, specifically: Acquire the microalgae culture solution image feature data in the target area, and simulate the environment through environmental simulation technology, add noise for simulation, enhance the effect of the microalgae culture solution image feature data, and enhance the microalgae culture solution image feature data; Acquire enhanced processed microalgae culture solution image feature data, construct a data training set according to the enhanced processed microalgae culture solution image feature data, and construct a microalgae growth prediction model based on a deep neural network; Inputting the data training set into the microalgae growth prediction model, using binary cross entropy as the loss function of the microalgae growth prediction model, and evaluating the microalgae growth prediction model using the binary cross entropy loss function; Convergence speed information is acquired, and when the convergence speed information is greater than preset convergence speed information, the microalgae growth prediction model is output.

[0006] Furthermore, in the microalgae feature fusion analysis method based on deep learning, the microalgae growth prediction model is optimized by generating adversarial networks, specifically: Acquiring information on the cell division law of microalgae under various environmental reproduction conditions, constructing microalgae biological constraints based on the information on the cell division law of microalgae under various environmental reproduction conditions, and training the microalgae growth prediction model based on the microalgae biological constraints; Through training, the generator generates initial microalgae growth prediction results based on the enhanced microalgae culture solution image feature data; Inputting the initial microalgae growth prediction result into the discriminator for judgment, and judging whether to accept the initial microalgae growth prediction result in combination with the microalgae biological constraint condition, and outputting the initial microalgae growth prediction result if the judgment result of the discriminator is acceptance; If the judgment result of the discriminator is not accepted, the generator continues to generate the next prediction result until the constraint conditions are met and the final prediction result is output, and the optimization of the microalgae growth prediction model is completed.

[0007] Furthermore, in the microalgae feature fusion analysis method based on deep learning, the microalgae culture solution image feature data in the target area is predicted by the microalgae growth prediction model to obtain the microalgae growth data within a preset time, specifically: Acquire the growth environment characteristics of microalgae in the target area, and count the microalgae data in the microalgae culture solution image feature data in the target area; Using the growth environment characteristics of the microalgae in the target area and the microalgae data in the microalgae culture solution image feature data in the target area as model input data; The model input data is input into the microalgae growth prediction model for prediction, a final prediction result when accepted by the discriminator is obtained, and the final prediction result when accepted by the discriminator is output as microalgae growth data within a preset time.

[0008] Furthermore, in the microalgae feature fusion analysis method based on deep learning, a microalgae growth and harvesting threshold range is set, and the microalgae harvesting time node is obtained according to the microalgae growth data within the preset time and the microalgae growth and harvesting threshold range, specifically: Setting a microalgae growth harvesting threshold range, and obtaining microalgae growth data of each timestamp from the microalgae growth data within the preset time, and determining whether the microalgae growth data of each timestamp is within the microalgae growth harvesting threshold range; The timestamp of the microalgae growth data within the microalgae growth harvesting threshold range is used as the microalgae harvesting time node, and the timestamp of the microalgae growth data not within the microalgae growth harvesting threshold range is used as the non-microalgae harvesting time node, and the microalgae harvesting time node is output.

[0009] Furthermore, in the microalgae feature fusion analysis method based on deep learning, pollution analysis is performed according to the real-time pollutant data at the current microalgae harvesting time node, and relevant control strategies are generated based on the pollution analysis results, specifically: Setting a pollution data threshold, and determining whether the real-time pollutant data at the current microalgae harvesting time node is greater than the pollution data threshold; When the real-time pollutant data at the current microalgae harvesting time node is greater than the pollution data threshold, a pollution search tag is constructed, and a search is performed based on the pollution search tag to obtain control measures related to pollution; A relevant control strategy is generated according to the pollution-related control measures, and the generated relevant control strategy is displayed in a preset manner.

[0010] A second aspect of the present invention provides a microalgae feature fusion analysis system based on deep learning, comprising a memory and a processor, wherein the memory comprises a microalgae feature fusion analysis method program based on deep learning, and when the microalgae feature fusion analysis method program based on deep learning is executed by the processor, the steps of any one of the microalgae feature fusion analysis methods based on deep learning are implemented.

[0011] The present invention solves the defects existing in the background technology and has the following beneficial effects: The present invention obtains the image feature data of microalgae culture solution in the target area, and constructs a microalgae growth prediction model based on a deep neural network, optimizes the microalgae growth prediction model through a generative adversarial network, and then predicts the image feature data of microalgae culture solution in the target area through the microalgae growth prediction model, obtains the microalgae growth data within a preset time, thereby setting the microalgae growth harvesting threshold range, obtains the microalgae harvesting time node according to the microalgae growth data within the preset time and the microalgae growth harvesting threshold range, and obtains the real-time pollutant data of the current microalgae harvesting time node, and finally performs pollution analysis according to the real-time pollutant data of the current microalgae harvesting time node, and generates relevant control strategies based on the pollution analysis results, and regulates according to the relevant control strategies. The present invention fuses the deep learning network and the generative adversarial network, and can fuse the cell division law of microalgae under different environments and the image features within the preset time to analyze the microalgae growth data in the target area, solves the pain point of image technology for microalgae growth recognition, and makes the harvesting of microalgae more reasonable. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, drawings of other embodiments can be obtained based on these drawings without paying creative work.

[0013] Figure 1 A portion of the microalgae training dataset is shown.

[0014] Figure 2 The overall flow chart of the microalgae feature fusion analysis method based on deep learning is shown; Figure 3 A partial flow chart of a microalgae feature fusion analysis method based on deep learning is shown; Figure 4 The system block diagram of the microalgae feature fusion analysis system based on deep learning is shown. DETAILED DESCRIPTION

[0015] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0016] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0017] like Figure 2 As shown, the first aspect of the present invention provides a microalgae feature fusion analysis method based on deep learning, comprising the following steps: S102: Acquire image feature data of microalgae culture solution in the target area, build a microalgae growth prediction model based on a deep neural network, and optimize the microalgae growth prediction model through a generative adversarial network; S104: predicting the image feature data of the microalgae culture solution in the target area by using the microalgae growth prediction model to obtain the microalgae growth data within a preset time; S106: setting a microalgae growth harvesting threshold range, obtaining a microalgae harvesting time node according to the microalgae growth data within a preset time and the microalgae growth harvesting threshold range, and obtaining real-time pollutant data at the current microalgae harvesting time node; S108: Perform pollution analysis based on the real-time pollutant data at the current microalgae harvesting time node, generate relevant control strategies based on the pollution analysis results, and perform control according to the relevant control strategies.

[0018] It should be noted that the present invention, by integrating deep learning networks and generative adversarial networks, can integrate the cell division rules of microalgae in different environments and image features within a preset time to analyze the microalgae growth data in the target area, thereby solving the pain points of image technology in identifying microalgae growth and making the harvesting of microalgae more reasonable.

[0019] Furthermore, in the microalgae feature fusion analysis method based on deep learning, the image feature data of the microalgae culture solution in the target area is obtained, and a microalgae growth prediction model is constructed based on a deep neural network, specifically: like Figure 1 As shown, a partial structural schematic diagram of microalgae and a processed image are shown, the image feature data of the microalgae culture solution in the target area is obtained, and the environment is simulated by environmental simulation technology, noise is added for simulation, the effect of the microalgae culture solution image feature data is enhanced, and the microalgae culture solution image feature data is enhanced; Exemplarily, the environmental simulation technology processing process includes: contrast adjustment, brightness adjustment, hue adjustment, grayscale transformation and other operations, noise simulation and removal including Gaussian noise removal, salt and pepper noise removal, etc., image enhancement including sharpening processing, Laplace algorithm processing, gamma transform processing, contrast-limited adaptive histogram equalization and other enhancement processing methods.

[0020] Acquire the image feature data of the enhanced microalgae culture solution, construct a data training set according to the image feature data of the enhanced microalgae culture solution, and construct a microalgae growth prediction model based on a deep neural network; It should be noted that deep neural networks include convolutional neural networks, recursive neural networks, etc.

[0021] The data training set is input into the microalgae growth prediction model, the binary cross entropy is used as the loss function of the microalgae growth prediction model, and the binary cross entropy loss function is used to evaluate the microalgae growth prediction model; The convergence speed information is obtained, and when the convergence speed information is greater than the preset convergence speed information, the microalgae growth prediction model is output.

[0022] like Figure 3 As shown, further, in the microalgae feature fusion analysis method based on deep learning, the microalgae growth prediction model is optimized by generating adversarial networks, specifically: S202: obtaining information on the cell division law of microalgae under various environmental reproduction conditions, constructing microalgae biological constraints based on the information on the cell division law of microalgae under various environmental reproduction conditions, and training a microalgae growth prediction model based on the microalgae biological constraints; It should be noted that the Generative Adversarial Network (GAN) is a deep learning model and one of the most promising methods for unsupervised learning on complex distributions in recent years. Generative adversarial networks can learn generative tasks in semi-supervised or unsupervised application scenarios. At present, generative adversarial networks have achieved amazing results in computer vision, natural language processing and other fields. Generative adversarial models are one of the most promising methods for unsupervised learning on complex data distributions in recent years. Generative Adversarial Network (GAN) is a generative model that learns by playing games between two neural networks. Generative adversarial networks can learn generative tasks without using labeled data. Generative adversarial networks consist of a generator and a discriminator. The generator randomly samples from the latent space as input, and its output needs to imitate the real samples in the training set as much as possible. The input of the discriminator is the real sample or the output of the generator, and its purpose is to distinguish the output of the generator from the real sample as much as possible. In this method, microalgae biological constraints are constructed based on the information of cell division rules of microalgae under various environmental reproduction conditions, and constraints are imposed through the constraints to optimize the training of the microalgae growth prediction model based on the microalgae biological constraints.

[0023] S204: Through training, the generator generates an initial microalgae growth prediction result according to the enhanced microalgae culture solution image feature data; S206: inputting the initial microalgae growth prediction result into the discriminator for judgment, and judging whether to accept the initial microalgae growth prediction result in combination with the microalgae biological constraint condition; if the judgment result of the discriminator is to accept, outputting the initial microalgae growth prediction result; S208: If the judgment result of the discriminator is not accepted, the generator continues to generate the next prediction result until the constraint conditions are met and the final prediction result is output, and the optimization of the microalgae growth prediction model is completed.

[0024] Exemplarily, the reproduction data includes data such as the volume ratio of microalgae per unit volume, the area ratio of microalgae per unit area, and the number information of microalgae per unit volume.

[0025] It should be noted that the cell division rules are different under different environmental reproduction conditions. Environmental reproduction conditions include temperature, humidity, salinity, chemical composition concentration data, etc. Due to the limitations of image technology, by integrating deep learning networks and generative adversarial networks, it is possible to integrate the cell division rules of microalgae in different environments and image features within a preset time to analyze the microalgae growth data in the target area, solving the pain points of image technology for microalgae growth identification and making the harvesting of microalgae more reasonable.

[0026] Furthermore, in the microalgae feature fusion analysis method based on deep learning, the microalgae culture solution image feature data in the target area is predicted by the microalgae growth prediction model to obtain the microalgae growth data within a preset time, specifically: Acquire the growth environment characteristics of microalgae in the target area, and count the microalgae data in the microalgae culture solution image feature data in the target area; The growth environment characteristics of the microalgae in the target area and the microalgae data in the image feature data of the microalgae culture solution in the target area are used as model input data; The model input data is input into the microalgae growth prediction model for prediction, and the final prediction result when the discriminator accepts it is obtained, and the final prediction result when the discriminator accepts it is output as the microalgae growth data within a preset time.

[0027] Furthermore, in the microalgae feature fusion analysis method based on deep learning, a microalgae growth and harvesting threshold range is set, and the microalgae harvesting time node is obtained according to the microalgae growth data within a preset time and the microalgae growth and harvesting threshold range, specifically: Setting a microalgae growth harvesting threshold range, and obtaining microalgae growth data of each timestamp from the microalgae growth data within a preset time, and determining whether the microalgae growth data of each timestamp is within the microalgae growth harvesting threshold range; The timestamp of the microalgae growth data within the microalgae growth harvesting threshold range is used as the microalgae harvesting time node, and the timestamp of the microalgae growth data not within the microalgae growth harvesting threshold range is used as the non-microalgae harvesting time node, and the microalgae harvesting time node is output.

[0028] It should be noted that the timestamp of the microalgae growth data within the microalgae growth harvesting threshold range is used as the microalgae harvesting time node, so that the harvesting time node meets the expected effect.

[0029] Furthermore, in the microalgae feature fusion analysis method based on deep learning, pollution analysis is performed according to the real-time pollutant data at the current microalgae harvesting time node, and relevant control strategies are generated based on the pollution analysis results, specifically: Set a pollution data threshold and determine whether the real-time pollutant data at the current microalgae harvesting time node is greater than the pollution data threshold; When the real-time pollutant data at the current microalgae harvesting time node is greater than the pollution data threshold, a pollution retrieval tag is constructed, and a search is performed based on the pollution retrieval tag to obtain pollution-related control measures; Relevant control strategies are generated according to the control measures related to pollution, and the generated relevant control strategies are displayed in a preset manner.

[0030] It should be noted that the pollution data include pollutant types, pollutant concentration data, etc. Through the present method, further determination can be made before harvesting, which can improve the rationality of microalgae harvesting.

[0031] In addition, the method also includes: obtaining microalgae surface charge characteristic data and working parameter characteristic data of the current electromagnetic device, and calculating the coupling characteristics of the magnetic field gradient and the microalgae surface charge based on the working parameter characteristic data of the current electromagnetic device and the microalgae surface charge characteristic data; introducing a particle swarm algorithm, setting the number of iterations based on the particle swarm algorithm, setting a coupling characteristic threshold, and judging whether the coupling characteristics of the magnetic field gradient and the microalgae surface charge are greater than the coupling characteristic threshold; when the coupling characteristics of the magnetic field gradient and the microalgae surface charge are greater than the coupling characteristic threshold, controlling the current electromagnetic device based on the working parameter characteristic data of the current electromagnetic device; when the coupling characteristics of the magnetic field gradient and the microalgae surface charge are not greater than the coupling characteristic threshold, iterating and adjusting the working parameter characteristic data of the current electromagnetic device based on the number of iterations until the coupling characteristics of the magnetic field gradient and the microalgae surface charge are greater than the coupling characteristic threshold.

[0032] It should be noted that due to different surface charges of microalgae, the electromagnetic equipment has different capture efficiencies when harvesting microalgae. The particle swarm algorithm is used to dynamically adjust the working parameter characteristic data (magnetic field gradient) of the current electromagnetic equipment according to the coupling characteristics of the magnetic field gradient and the surface charge of microalgae, thereby improving the capture efficiency of microalgae harvesting. Among them, the coupling characteristics of the magnetic field gradient and the surface charge of microalgae refer to the phenomenon that two or more objects interact with each other through the action of the magnetic field. In the process of microalgae separation, the magnetic field gradient can effectively affect the movement trajectory of microalgae cells, thereby improving the separation efficiency.

[0033] like Figure 4 As shown, the second aspect of the present invention provides a microalgae feature fusion analysis system 4 based on deep learning, including a memory 41 and a processor 42. The memory 41 includes a microalgae feature fusion analysis method program based on deep learning. When the microalgae feature fusion analysis method program based on deep learning is executed by the processor 42, the technical solution of the above-mentioned microalgae feature fusion analysis method based on deep learning is implemented.

[0034] In summary, the present invention obtains the image feature data of the microalgae culture solution in the target area, and constructs a microalgae growth prediction model based on a deep neural network, optimizes the microalgae growth prediction model through a generative adversarial network, and then predicts the image feature data of the microalgae culture solution in the target area through the microalgae growth prediction model, obtains the microalgae growth data within the preset time, thereby setting the microalgae growth harvesting threshold range, obtains the microalgae harvesting time node according to the microalgae growth data within the preset time and the microalgae growth harvesting threshold range, and obtains the real-time pollutant data of the current microalgae harvesting time node, and finally performs pollution analysis according to the real-time pollutant data of the current microalgae harvesting time node, and generates relevant control strategies based on the pollution analysis results, and regulates according to the relevant control strategies. The present invention is integrated by integrating the deep learning network and the generative adversarial network, and can integrate the cell division law of microalgae under different environments and the image features within the preset time to analyze the microalgae growth data in the target area, solves the pain point of image technology for microalgae growth recognition, and makes the harvesting of microalgae more reasonable.

[0035] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0036] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0037] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0038] A person of ordinary skill in the art can understand that: all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: a mobile storage device, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0039] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the methods of each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0040] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A microalgae feature fusion analysis method based on deep learning, characterized in that: The following steps are involved: Acquire image feature data of microalgae culture solution in the target area, build a microalgae growth prediction model based on a deep neural network, and optimize the microalgae growth prediction model through a generative adversarial network; Predicting the image feature data of the microalgae culture solution in the target area by using the microalgae growth prediction model to obtain microalgae growth data within a preset time; Setting a microalgae growth and harvesting threshold range, obtaining a microalgae harvesting time node according to the microalgae growth data within the preset time and the microalgae growth and harvesting threshold range, and obtaining real-time pollutant data at the current microalgae harvesting time node; Pollution analysis is performed according to the real-time pollutant data at the current microalgae harvesting time node, and relevant control strategies are generated based on the pollution analysis results, and control is performed according to the relevant control strategies.

2. The method for fusion analysis of microalgae features based on deep learning according to claim 1, characterized in that: The image feature data of the microalgae culture solution in the target area is obtained, and a microalgae growth prediction model is constructed based on a deep neural network. Specifically: Acquire the microalgae culture solution image feature data in the target area, and simulate the environment through environmental simulation technology, add noise for simulation, enhance the effect of the microalgae culture solution image feature data, and enhance the microalgae culture solution image feature data; Acquire enhanced processed microalgae culture solution image feature data, construct a data training set according to the enhanced processed microalgae culture solution image feature data, and construct a microalgae growth prediction model based on a deep neural network; Inputting the data training set into the microalgae growth prediction model, using binary cross entropy as the loss function of the microalgae growth prediction model, and evaluating the microalgae growth prediction model using the binary cross entropy loss function; Convergence speed information is acquired, and when the convergence speed information is greater than preset convergence speed information, the microalgae growth prediction model is output.

3. The microalgae feature fusion analysis method based on deep learning according to claim 1, characterized in that: The microalgae growth prediction model is optimized by generating adversarial networks, specifically: Acquiring information on the cell division law of microalgae under various environmental reproduction conditions, constructing microalgae biological constraints based on the information on the cell division law of microalgae under various environmental reproduction conditions, and training the microalgae growth prediction model based on the microalgae biological constraints; Through training, the generator generates initial microalgae growth prediction results based on the enhanced microalgae culture solution image feature data; Inputting the initial microalgae growth prediction result into a discriminator for judgment, and judging whether to accept the initial microalgae growth prediction result in combination with the microalgae biological constraint condition, and outputting the initial microalgae growth prediction result if the judgment result of the discriminator is acceptance; If the judgment result of the discriminator is not accepted, the generator continues to generate the next prediction result until the constraint conditions are met and the final prediction result is output, and the optimization of the microalgae growth prediction model is completed.

4. The method for fusion analysis of microalgae features based on deep learning according to claim 1, characterized in that: The microalgae growth prediction model is used to predict the microalgae culture solution image feature data in the target area to obtain microalgae growth data within a preset time, specifically: Acquire the growth environment characteristics of microalgae in the target area, and count the microalgae data in the microalgae culture solution image feature data in the target area; Using the growth environment characteristics of the microalgae in the target area and the microalgae data in the microalgae culture solution image feature data in the target area as model input data; The model input data is input into the microalgae growth prediction model for prediction, a final prediction result when accepted by the discriminator is obtained, and the final prediction result when accepted by the discriminator is output as microalgae growth data within a preset time.

5. The method for fusion analysis of microalgae features based on deep learning according to claim 1, characterized in that: The microalgae growth and harvesting threshold range is set, and the microalgae harvesting time node is obtained according to the microalgae growth data within the preset time and the microalgae growth and harvesting threshold range, specifically: Setting a microalgae growth harvesting threshold range, and obtaining microalgae growth data of each timestamp from the microalgae growth data within the preset time, and determining whether the microalgae growth data of each timestamp is within the microalgae growth harvesting threshold range; The timestamp of the microalgae growth data within the microalgae growth harvesting threshold range is used as the microalgae harvesting time node, and the timestamp of the microalgae growth data not within the microalgae growth harvesting threshold range is used as the non-microalgae harvesting time node, and the microalgae harvesting time node is output.

6. The method for fusion analysis of microalgae features based on deep learning according to claim 1, characterized in that: Pollution analysis is performed according to the real-time pollutant data at the current microalgae harvesting time node, and relevant control strategies are generated based on the pollution analysis results, specifically: Setting a pollution data threshold, and determining whether the real-time pollutant data at the current microalgae harvesting time node is greater than the pollution data threshold; When the real-time pollutant data at the current microalgae harvesting time node is greater than the pollution data threshold, a pollution search tag is constructed, and a search is performed based on the pollution search tag to obtain control measures related to pollution; A relevant control strategy is generated according to the pollution-related control measures, and the generated relevant control strategy is displayed in a preset manner.

7. A microalgae feature fusion analysis system based on deep learning, characterized in that: It comprises a memory and a processor, wherein the memory comprises a microalgae feature fusion analysis method program based on deep learning, and when the microalgae feature fusion analysis method program based on deep learning is executed by the processor, the steps of the microalgae feature fusion analysis method based on deep learning as described in any one of claims 1 to 6 are implemented.

Images