A non-destructive testing method and device for multi-variety tomatoes

Abstract

The present invention provides a non-destructive detection method and device for multi-variety tomatoes, belonging to the technical field of agricultural product inspection and testing. The method includes: inputting the spectral data of each tomato variety into the soluble solid content detection model corresponding to each tomato variety to obtain the predicted value of the soluble solid content corresponding to each tomato variety; determining the target tomato variety from all tomato varieties according to the observed value and the predicted value of the soluble solid content of each tomato variety, and using the soluble solid content detection model of the target tomato variety as the target soluble solid content detection model; and performing non-destructive detection on the soluble solid content of any tomato variety by using the target soluble solid content detection model. By establishing the soluble solid content detection model for each tomato variety and determining the target soluble solid content detection model with the best performance therefrom to perform non-destructive detection on any tomato variety, the present invention realizes accurate and rapid non-destructive detection of multi-variety tomatoes.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural product inspection and testing, and particularly relates to a non-destructive detection method and device for multi-variety tomatoes. Background Art

[0002] Tomatoes, as extremely important vegetable consumer goods worldwide, have good edible and processing values.

[0003] However, due to the large variety of tomatoes, even if the soluble solid contents of some tomato varieties are very close, the differences in their shapes, sizes, and internal structures still result in the lack of a unified and effective non-destructive detection method.

[0004] How to accurately and quickly evaluate the soluble solid quality of multi-variety tomatoes has important theoretical significance and practical value for improving China's vegetable detection ability, ensuring the quality of China's vegetables and related deep-processed products, and enhancing the international market competitiveness. Summary of the Invention

[0005] The present invention provides a non-destructive detection method and device for multi-variety tomatoes to solve the defects in the prior art and achieve accurate detection of the soluble solids of tomatoes.

[0006] In a first aspect, the present invention provides a non-destructive detection method for multi-variety tomatoes, including: inputting the spectral data of each tomato variety into the soluble solid detection model corresponding to each tomato variety to obtain the predicted value of the soluble solid corresponding to each tomato variety; determining the target tomato variety from all tomato varieties according to the observed value and predicted value of the soluble solid of each tomato variety, and using the soluble solid detection model of the target tomato variety as the target soluble solid detection model; and performing non-destructive detection on the soluble solid of any tomato variety by using the target soluble solid detection model.

[0007] According to the non-destructive detection method for multi-variety tomatoes provided by the present invention, after using the soluble solid detection model of the target tomato variety as the target soluble solid detection model, it further includes:

[0008] Step 201: Inputting the spectral data of any non-target tomato variety into the spectral curve correction network to obtain the corrected spectral feature data of any non-target tomato variety;

[0009] Step 202: Inputting the corrected spectral feature data into the target soluble solid detection model to obtain the predicted value of the soluble solid of any non-target tomato variety;

[0010] Step 203: Based on the predicted value and the observed value of the soluble solids of any non-target tomato variety, adjust the spectral curve correction network;

[0011] Step 204: Iteratively execute Step 201 to Step 203 until a preset iteration stop condition is met, and use the finally adjusted spectral curve correction network as the target spectral curve correction network.

[0012] According to a non-destructive detection method for multi-variety tomatoes provided by the present invention, before inputting the spectral data of each tomato variety into the soluble solids detection model corresponding to each tomato variety, it further includes: obtaining the spectral data of each tomato variety and the observed value of the soluble solids corresponding to each tomato variety; establishing a soluble solids detection model for each tomato variety according to the spectral data and the observed value of the soluble solids.

[0013] According to a non-destructive detection method for multi-variety tomatoes provided by the present invention, determining a target tomato variety from all tomato varieties according to the observed value and the predicted value of the soluble solids of each tomato variety includes: determining a target tomato variety from all tomato varieties according to the correlation coefficient and the root mean square error between the observed value and the predicted value of the soluble solids.

[0014] According to a non-destructive detection method for multi-variety tomatoes provided by the present invention, the non-destructive detection of the soluble solids of any tomato variety by using the target soluble solids detection model includes:

[0015] Obtain the type of any tomato variety;

[0016] When it is determined that the any tomato variety is a target tomato variety, perform non-destructive detection on the target tomato variety by using the target soluble solids detection model;

[0017] When it is determined that the any tomato variety is any non-target tomato variety, perform non-destructive detection on the any non-target tomato variety by using the target spectral curve correction network and the target soluble solids detection model.

[0018] According to a non-destructive detection method for multi-variety tomatoes provided by the present invention, the spectral data is near-infrared spectral data; both the soluble solids detection model and the spectral curve correction network are established based on a convolutional neural network.

[0019] In a second aspect, the present invention further provides a non-destructive detection device for multi-variety tomatoes, including:

[0020] The first module is configured to input the spectral data of each tomato variety into the soluble solid content detection model corresponding to each tomato variety, and obtain the predicted value of the soluble solid content corresponding to each tomato variety;

[0021] The second module is configured to determine a target tomato variety from all tomato varieties according to the observed value and the predicted value of the soluble solid content of each tomato variety, and use the soluble solid content detection model of the target tomato variety as the target soluble solid content detection model;

[0022] The third module is configured to perform non-destructive detection of the soluble solid content of any tomato variety by using the target soluble solid content detection model.

[0023] In a third aspect, the present invention provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the multi-variety tomato non-destructive detection method as described in any one of the above is implemented.

[0024] In a fourth aspect, the present invention further provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the multi-variety tomato non-destructive detection method as described in any one of the above is implemented.

[0025] In a fifth aspect, the present invention further provides a computer program product, including a computer program. When the computer program is executed by a processor, the multi-variety tomato non-destructive detection method as described in any one of the above is implemented.

[0026] The multi-variety tomato non-destructive detection method and device provided by the present invention establish a soluble solid content detection model for each tomato variety, and determine the best-performing target soluble solid content detection model from them to perform non-destructive detection of the soluble solid content of any tomato variety, realizing accurate and rapid non-destructive detection of multi-variety tomatoes. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0028] Figure 1 is one of the flow diagrams of the multi-variety tomato non-destructive detection method provided by the present invention;

[0029] Figure 2 is a schematic diagram of the average spectral curves of two tomato varieties provided by the present invention;

[0030] Figure 3 It is the second flow schematic diagram of the non-destructive detection method for multi-variety tomatoes provided by the present invention;

[0031] Figure 4 It is one of the schematic diagrams of the original spectral curve and the corrected characteristic curve provided by the present invention;

[0032] Figure 5 It is the schematic diagram of the predicted values of soluble solids before and after calibration for Variety 2 provided by the present invention;

[0033] Figure 6 It is the second schematic diagram of the original spectral curve and the corrected characteristic curve provided by the present invention;

[0034] Figure 7 It is the schematic diagram of the predicted values of soluble solids before and after calibration for Variety 3 provided by the present invention;

[0035] Figure 8 It is the structural schematic diagram of the non-destructive detection device for multi-variety tomatoes provided by the present invention;

[0036] Figure 9 It is the structural schematic diagram of the electronic device provided by the present invention. Detailed implementation manners

[0037] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below with reference to the accompanying drawings in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present invention without making creative efforts shall fall within the protection scope of the present invention.

[0038] It should be noted that in the description of the embodiments of the present invention, the terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements but also includes other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "including one..." does not exclude the presence of additional identical elements in the process, method, article or device including the said element. The orientation or positional relationship indicated by terms such as "upper", "lower", etc. is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as a limitation on the present invention. Unless otherwise expressly specified and limited, the terms "mount", "connect", "couple" should be understood in a broad sense. For example, it may be a fixed connection, a detachable connection, or an integral connection; it may be a mechanical connection or an electrical connection; it may be directly connected or indirectly connected through an intermediate medium, and it may be the internal communication of two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0039] The terms "first", "second", etc. in this application are used to distinguish similar objects and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances so that the embodiments of this application can be implemented in an order other than those illustrated or described here, and the objects distinguished by "first", "second", etc. are generally of the same category, and the number of objects is not limited. For example, the first object can be one or multiple. In addition, "and / or" means at least one of the connected objects, and the character " / ", generally indicates that the related objects before and after are in an "or" relationship.

[0040] Next, in combination with Figures 1-9 describe the multi-variety tomato non-destructive detection method and device provided by the embodiments of the present invention.

[0041] Figure 1 is one of the flow diagrams of the multi-variety tomato non-destructive detection method provided by the present invention, as Figure 1 shown, including but not limited to the following steps:

[0042] Step 101: Input the spectral data of each tomato variety into the soluble solid content detection model corresponding to each tomato variety, and obtain the predicted value of the soluble solid content corresponding to each tomato variety.

[0043] Optionally, the spectral data is near-infrared spectral data, with the wavelength range for collection being 550 - 1100 nm and the band interval being 0.265 nm. The spectral data can be collected using a near-infrared online collection system.

[0044] Optionally, the soluble solids detection model is a pre-established convolutional neural network model.

[0045] The present invention can use the spectral data of each tomato variety as input and the observed values of soluble solids corresponding to each tomato variety as output to train the soluble solids detection model.

[0046] It should be noted that the observed values of soluble solids for each tomato variety can be obtained by juicing the whole fruit and measuring the tomato juice filtered through gauze using a handheld refractometer.

[0047] Step 102: Determine the target tomato variety from all tomato varieties based on the observed values and predicted values of soluble solids for each tomato variety, and use the soluble solids detection model of the target tomato variety as the target soluble solids detection model.

[0048] The present invention can evaluate the soluble solids detection model corresponding to each tomato variety based on the degree of similarity between the observed values and predicted values of soluble solids for each tomato variety.

[0049] Based on the evaluation results, the tomato variety with the best evaluation can be determined from all tomato varieties as the target tomato variety, and the soluble solids detection model corresponding to the target tomato variety can be used as the target soluble solids detection model.

[0050] Step 103: Use the target soluble solids detection model to perform non-destructive detection of the soluble solids of any tomato variety.

[0051] For example, in the embodiment of the present invention, a total of 3 soluble solids detection models for tomato varieties are established, and a target soluble solids detection model with the best performance can be determined from the 3 soluble solids detection models. And use the target soluble solids detection model to perform non-destructive detection of the soluble solids of any tomato variety.

[0052] It should be noted that any tomato variety in the present invention can be one of the above 3 tomato varieties or other tomato varieties.

[0053] The multi-variety tomato non-destructive detection method provided by the present invention realizes accurate and rapid non-destructive detection of multi-variety tomatoes by establishing soluble solids detection models for each tomato variety and determining the target soluble solids detection model with the best performance from them to perform non-destructive detection of the soluble solids of any tomato variety.

[0054] Based on the content of the above embodiments, as an alternative embodiment, the non-destructive detection method for multi-variety tomatoes provided by the present invention further includes, after using the soluble solid content detection model of the target tomato variety as the target soluble solid content detection model:

[0055] Step 201: Input the spectral data of any non-target tomato variety into the spectral curve correction network to obtain the corrected spectral feature data of the any non-target tomato variety;

[0056] Step 202: Input the corrected spectral feature data into the target soluble solid content detection model to obtain the predicted value of the soluble solid content of the any non-target tomato variety;

[0057] Step 203: Adjust the spectral curve correction network based on the predicted value and the observed value of the soluble solid content of the any non-target tomato variety;

[0058] Step 204: Iteratively execute Steps 201 to 203 until a preset iteration stop condition is met, and use the finally adjusted spectral curve correction network as the target spectral curve correction network.

[0059] Wherein, any non-target tomato variety is any tomato variety that does not belong to the target tomato variety; the spectral curve correction network is also established based on a convolutional neural network.

[0060] Specifically, in Step 203, the loss function of the spectral curve correction network can be determined according to the predicted value and the observed value of the soluble solid content, and then the spectral curve correction network can be adjusted according to the loss function. Correspondingly, the preset iteration stop condition in Step 204 can be set such that the loss function obtains the global minimum value.

[0061] The spectral curve correction network can extract features from the spectral data of any non-target tomato variety and output the corrected spectral feature data.

[0062] Then, input the corrected spectral feature data into the target soluble solid content detection model, and iteratively adjust the parameters in the spectral curve correction network according to the predicted value and the observed value of the soluble solid content.

[0063] Wherein, the preset iteration stop condition can be a preset number of iterations, such as 100 times; the preset iteration stop condition can also be set according to the error between the predicted value and the observed value of the soluble solid content. For example, when the root mean square error between the two is less than a preset threshold, the iteration stops.

[0064] By continuously iteratively adjusting the spectral curve correction network, after reaching the preset iteration stop condition, the adjusted spectral curve correction network is used as the target spectral curve correction network, and the target spectral curve correction network and the target soluble solid content detection model are used to perform non-destructive detection on any non-target tomato variety.

[0065] Based on the content of the above embodiments, as an optional embodiment, the multi-variety tomato non-destructive detection method provided by the present invention, the non-destructive detection of the soluble solid content of any tomato variety by using the target soluble solid content detection model includes: obtaining the type of any tomato variety; in the case of determining that the any tomato variety is a target tomato variety, using the target soluble solid content detection model to perform non-destructive detection on the target tomato variety; in the case of determining that the any tomato variety is any non-target tomato variety, using the target spectral curve correction network and the target soluble solid content detection model to perform non-destructive detection on the any non-target tomato variety.

[0066] For the target tomato variety, the present invention can directly use the target soluble solid content detection model to perform non-destructive detection; for any non-target tomato variety, it is necessary to first train the spectral curve correction network to obtain the target spectral curve correction network, and then perform non-destructive detection based on the target spectral curve correction network and the target soluble solid content detection model.

[0067] The present invention organically combines spectral technology with deep learning algorithms to establish a general and rapid detection method for the soluble solid content of tomato samples of different varieties, overcoming the inherent technical defect that it is difficult to perform general and rapid detection due to the large spectral line differences among tomato samples of different varieties with similar soluble solid contents. Compared with the traditional method of separately modeling and predicting for different varieties, the present invention has the following advantages:

[0068] The present invention successfully solves the problem of a large number of tomato varieties and the need for individual modeling and verification one by one, thus saving a large amount of manpower and material resources; this method is simple, fast, non-destructive, and reliable, and does not require separately establishing and verifying models for each tomato variety in each region one by one. The operator does not need to have professional knowledge and the samples do not require specific pretreatment; the technology of the present invention can be directly applied to the front line of production and sales, meeting the needs of producers and consumers.

[0069] The present invention completes the correction of the spectral curves of random variety tomato samples to the standard soluble solid content spectral curves through a small number of typical samples, and realizes the general and rapid non-destructive detection of the soluble solid content quality of tomato samples of different varieties through visible / near-infrared spectral technology.

[0070] The present invention is applicable to on-line production, and solves the problems that current tomato detection is carried out in a single region, for a single variety, with separate modeling and verification, resulting in poor model generality and being time-consuming and laborious, making it difficult to meet the requirements of on-line sorting of bulk tomatoes.

[0071] To better illustrate the non-destructive detection method for multiple varieties of tomatoes provided by the present invention, the following specific embodiment is taken as an example for illustration. This example is only for the explanation of the embodiments of the present invention and is not regarded as a limitation on the protection scope of the embodiments of the present invention.

[0072] Experimental materials: A total of 1800 fresh tomato samples of two varieties (variety 1 and variety 2, 900 for each variety) were picked in the field and transported back to the laboratory in fruit baskets.

[0073] Near-infrared diffuse transmission spectral data of all 1800 samples were collected, and the quality index of soluble solids content (SSC) corresponding to all 1800 samples was measured as the observed value of soluble solids content. Figure 2 It is a schematic diagram of the average spectral curves of two tomato varieties provided by the present invention, where the abscissa represents the wavelength and the ordinate represents the intensity.

[0074] Secondly, the spectral data of tomato samples of each variety were randomly assigned at a ratio of 2:1. For each tomato variety, a calibration set (including 600 tomatoes) and a prediction set (including 300 tomatoes) were obtained respectively. Using a self-built convolutional neural network, the correlation between the spectral data of the calibration set of each tomato variety and the observed value of soluble solids content was established respectively, that is, the soluble solids content detection model.

[0075] It should be noted that the convolutional neural network in the present invention can be built using the relevant modules provided in Python software.

[0076] For tomato variety 1 in the calibration set, the correlation coefficient (abbreviation Rp) between the observed value and the predicted value of soluble solids content was 0.91, and the root mean square error (abbreviation RMSEp) was 0.26; for tomato variety 2 in the calibration set, Rp was 0.74 and RMSEp was 0.39.

[0077] The spectral data of the prediction sets of variety 1 and variety 2 were respectively imported into their corresponding soluble solids content detection models, and the following results could be obtained: the correlation coefficient Rp of variety 1 was 0.85, and the correlation coefficient Rp of variety 2 was 0.69; the root mean square error of variety 1 was 0.27; the root mean square error of variety 2 was 0.46.

[0078] The above results show that, compared with Variety 2, the convolutional neural network (i.e., the soluble solids detection model) built for Variety 1 has better prediction accuracy. Therefore, the target soluble solids detection model in the present invention is the soluble solids detection model of Variety 1.

[0079] Figure 3 It is the second flow schematic diagram of the non-destructive detection method for multi-variety tomatoes provided by the present invention. As Figure 3 shown, after determining that the convolutional neural network model (i.e., the deep learning network) of Variety 1 is the target soluble solids detection model, the weights of the second middle layer and the output layer of the convolutional neural network model of Variety 1 can be frozen.

[0080] It should be noted that the first middle layer and the second middle layer both belong to the hidden layers of the convolutional neural network, where Conv represents the convolutional layer and the Normalization layer is the ReLU layer.

[0081] Input the spectral data of Variety 2 samples into a separate spectral curve correction network. Further, connect the spectral curve correction network to the second middle layer and the subsequent network of the built convolutional neural network of Variety 1, and train the spectral curve correction network according to the Variety 2 samples until the input results of the convolutional neural network model (deep learning network) meet the pre-set conditions, and finally obtain the target spectral curve correction network and the corresponding corrected spectral feature data, that is, the correction of the spectral data is completed.

[0082] Figure 4 It is one of the schematic diagrams of the original spectral curve and the corrected characteristic curve provided by the present invention. The original spectral curve is established based on the spectral data, and the corrected characteristic curve is established based on the corrected spectral feature data. Combining Figure 3 and Figure 4 it can be seen that after the spectral curve correction network and the second middle layer of the deep learning network extract the features of the spectral data of Variety 2, the corrected spectral curve of Variety 2 obtained can also well reflect the characteristics of the original spectral curve. It should be noted that the corrected spectral feature data is actually the data obtained after extracting the features from the spectral data.

[0083] In this embodiment, the target spectral curve correction network and the convolutional neural network of Variety 1 can be used to predict the soluble solids of Variety 2 to obtain the soluble solids prediction value.

[0084] Figure 5 It is the schematic diagram of the soluble solids prediction values of Variety 2 before and after correction provided by the present invention. As Figure 5As shown in the figure, the schematic diagram of the predicted value corresponding to Variety 2 before calibration is obtained based on the soluble solid content detection model of Variety 2; the schematic diagram of the predicted value corresponding to Variety 2 after calibration is obtained based on the target spectral curve calibration network and the convolutional neural network of Variety 1 (target spectral curve calibration network). From Figure 5 It can be seen that the prediction effect of the soluble solid content of Variety 2 is improved from Rp = 0.69 and RMSEp = 0.46 to Rp = 0.82 and RMSEp = 0.34. Among them, the measured value of SSC is the observed value of soluble solid content; the predicted value of SSC is the predicted value of soluble solid content.

[0085] Finally, 900 tomatoes of Variety 3 were separately collected for verification. The calibration set and prediction set were still divided according to the ratio of 2:1, and the soluble solid content detection model of Variety 3 was established. And the soluble solid content detection model of Variety 1 was still the target soluble solid content detection model.

[0086] Figure 6 This is the second schematic diagram of the original spectral curve and the calibrated characteristic curve provided by the present invention. From Figure 6 It can be seen that for Variety 3, the calibrated characteristic curve obtained based on the calibrated spectral characteristic data can also well reflect the characteristics of the original spectral curve.

[0087] Figure 7 This is the schematic diagram of the predicted value of the soluble solid content of Variety 3 before and after calibration provided by the present invention. As Figure 7 shown, the schematic diagram of the predicted value corresponding to Variety 3 before calibration is obtained based on the soluble solid content detection model of Variety 3; the schematic diagram of the predicted value corresponding to Variety 3 after calibration is obtained based on the target spectral curve calibration network (established based on Variety 3) and the convolutional neural network of Variety 1 (target spectral curve calibration network). Among them, the measured value of SSC is the observed value of soluble solid content; the predicted value of SSC is the predicted value of soluble solid content.

[0088] As Figure 7 shown, the correlation coefficient and root mean square error of the soluble solid content detection model established based on Variety 3 are Rp = 0.78 and RMSEp = 0.28 respectively. However, after migrating and using the soluble solid content detection model of Variety 1, the prediction accuracy is improved to Rp = 0.82 and RMSEp = 0.26.

[0089] It can be seen from this that the present invention can effectively calibrate the spectral curves of different varieties of tomatoes. The finally applied general neural network of Variety 1 has good accuracy and wide applicability.

[0090] Figure 8 This is the structural schematic diagram of the non-destructive detection device for multi-variety tomatoes provided by the present invention. As Figure 8As shown, the system includes: a first module 801, a second module 802, and a third module 803.

[0091] The first module 801 is configured to input the spectral data of each tomato variety into the soluble solid content detection model corresponding to each tomato variety, and obtain the predicted value of the soluble solid content corresponding to each tomato variety.

[0092] The second module 802 is configured to determine a target tomato variety from all tomato varieties according to the observed value and the predicted value of the soluble solid content of each tomato variety, and use the soluble solid content detection model of the target tomato variety as the target soluble solid content detection model.

[0093] The third module 803 is configured to perform non-destructive detection of the soluble solid content of any tomato variety by using the target soluble solid content detection model.

[0094] The multi-variety tomato non-destructive detection device provided by the present invention realizes accurate and rapid non-destructive detection of multi-variety tomatoes by establishing a soluble solid content detection model for each tomato variety, and determining the best-performing target soluble solid content detection model from them to perform non-destructive detection of the soluble solid content of any tomato variety.

[0095] It should be noted that the multi-variety tomato non-destructive detection device provided by the embodiments of the present invention can execute the multi-variety tomato non-destructive detection method described in any of the above embodiments during specific operation, and this embodiment will not be elaborated herein.

[0096] Figure 9 is a schematic structural diagram of an electronic device provided by the present invention. As Figure 9 shown, the electronic device may include: a processor 910, a communication interface 920, a memory 930, and a communication bus 940. Among them, the processor 910, the communication interface 920, and the memory 930 communicate with each other through the communication bus 940. The processor 910 can call the logical instructions in the memory 930 to execute the multi-variety tomato non-destructive detection method, which includes: inputting the spectral data of each tomato variety into the soluble solid content detection model corresponding to each tomato variety, and obtaining the predicted value of the soluble solid content corresponding to each tomato variety; determining a target tomato variety from all tomato varieties according to the observed value and the predicted value of the soluble solid content of each tomato variety, and using the soluble solid content detection model of the target tomato variety as the target soluble solid content detection model; performing non-destructive detection of the soluble solid content of any tomato variety by using the target soluble solid content detection model.

[0097] In addition, when the logical instructions in the above-mentioned memory 930 are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0098] On the other hand, the present invention also provides a computer program product. The computer program product includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the non-destructive detection method for multi-variety tomatoes provided by the above-mentioned various methods. The method includes: inputting the spectral data of each tomato variety into the soluble solid content detection model corresponding to each tomato variety to obtain the predicted value of the soluble solid content corresponding to each tomato variety; determining a target tomato variety from all tomato varieties according to the observed value and the predicted value of the soluble solid content of each tomato variety, and using the soluble solid content detection model of the target tomato variety as the target soluble solid content detection model; and using the target soluble solid content detection model to perform non-destructive detection on the soluble solid content of any tomato variety.

[0099] On another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it is used to execute the non-destructive detection method for multi-variety tomatoes provided by the above-mentioned various embodiments. The method includes: inputting the spectral data of each tomato variety into the soluble solid content detection model corresponding to each tomato variety to obtain the predicted value of the soluble solid content corresponding to each tomato variety; determining a target tomato variety from all tomato varieties according to the observed value and the predicted value of the soluble solid content of each tomato variety, and using the soluble solid content detection model of the target tomato variety as the target soluble solid content detection model; and using the target soluble solid content detection model to perform non-destructive detection on the soluble solid content of any tomato variety.

[0100] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0101] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0102] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A non-destructive testing method for multi-variety tomatoes, characterized in that, Including: Step 101: Input the spectral data of each tomato variety into the soluble solid content detection model corresponding to each tomato variety to obtain the predicted value of the soluble solid content corresponding to each tomato variety; Step 102: Determine the target tomato variety from all tomato varieties according to the observed value and predicted value of the soluble solid content of each tomato variety, and use the soluble solid content detection model of the target tomato variety as the target soluble solid content detection model. After using the soluble solid content detection model of the target tomato variety as the target soluble solid content detection model, it further includes: Step 201: Input the spectral data of any non-target tomato variety into the spectral curve correction network to obtain the corrected spectral feature data of any non-target tomato variety; Step 202: Input the corrected spectral feature data into the target soluble solid content detection model to obtain the predicted value of the soluble solid content of any non-target tomato variety; Step 203: Adjust the spectral curve correction network based on the predicted value and observed value of the soluble solid content of any non-target tomato variety; Step 204: Iteratively execute steps 201 to 203 until the preset iteration stop condition is met, and use the finally adjusted spectral curve correction network as the target spectral curve correction network; Step 103: Use the target soluble solid content detection model to perform non-destructive detection on the soluble solid content of any tomato variety; The determining the target tomato variety from all tomato varieties according to the observed value and predicted value of the soluble solid content of each tomato variety includes: Determine the target tomato variety from all tomato varieties according to the correlation coefficient and root mean square error between the observed value and predicted value of the soluble solid content; The using the target soluble solid content detection model to perform non-destructive detection on the soluble solid content of any tomato variety includes: Obtain the type of any tomato variety; When it is determined that the any tomato variety is the target tomato variety, use the target soluble solid content detection model to perform non-destructive detection on the target tomato variety; When it is determined that the any tomato variety is any non-target tomato variety, use the target spectral curve correction network and the target soluble solid content detection model to perform non-destructive detection on the any non-target tomato variety.

2. The non-destructive detection method for multi-variety tomatoes according to claim 1, characterized in that, Before inputting the spectral data of each tomato variety into the soluble solid content detection model corresponding to each tomato variety, it further includes: Obtain the spectral data of each tomato variety and the observed value of the soluble solid content corresponding to each tomato variety; According to the spectral data and the observed value of the soluble solid content, establish the soluble solid content detection model of each tomato variety.

3. The non-destructive detection method for multi-variety tomatoes according to claim 1, characterized in that The spectral data is near-infrared spectral data; Both the soluble solid content detection model and the spectral curve correction network are established based on a convolutional neural network.

4. A non-destructive testing device for multi-variety tomatoes, characterized in that, Including: The first module is used to execute step 101: input the spectral data of each tomato variety into the soluble solid content detection model corresponding to each tomato variety, and obtain the predicted value of the soluble solid content corresponding to each tomato variety; The second module is used to execute step 102: determine the target tomato variety from all tomato varieties according to the observed value and predicted value of the soluble solid content of each tomato variety, and use the soluble solid content detection model of the target tomato variety as the target soluble solid content detection model. After using the soluble solid content detection model of the target tomato variety as the target soluble solid content detection model, it further includes: Step 201: input the spectral data of any non-target tomato variety into the spectral curve correction network, and obtain the corrected spectral feature data of any non-target tomato variety; Step 202: input the corrected spectral feature data into the target soluble solid content detection model, and obtain the predicted value of the soluble solid content of any non-target tomato variety; Step 203: adjust the spectral curve correction network based on the predicted value and observed value of the soluble solid content of any non-target tomato variety; Step 204: iteratively execute steps 201 to 203 until a preset iteration stop condition is met, and use the finally adjusted spectral curve correction network as the target spectral curve correction network; The third module is used to execute step 103: use the target soluble solid content detection model to perform non-destructive detection on the soluble solid content of any tomato variety; The determining the target tomato variety from all tomato varieties according to the observed value and predicted value of the soluble solid content of each tomato variety includes: determining the target tomato variety from all tomato varieties according to the correlation coefficient and root mean square error between the observed value and predicted value of the soluble solid content; The using the target soluble solid content detection model to perform non-destructive detection on the soluble solid content of any tomato variety includes: obtaining the type of any tomato variety; when it is determined that the any tomato variety is the target tomato variety, using the target soluble solid content detection model to perform non-destructive detection on the target tomato variety; when it is determined that the any tomato variety is any non-target tomato variety, using the target spectral curve correction network and the target soluble solid content detection model to perform non-destructive detection on any non-target tomato variety.

5. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the multi-variety tomato non-destructive detection method according to any one of claims 1 to 3.

6. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the multi-variety tomato non-destructive detection method according to any one of claims 1 to 3.

7. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the multi-variety tomato non-destructive detection method according to any one of claims 1 to 3.

Images