Pulp quality detector, pulp quality detection method and computer program product

Abstract

The invention discloses a pulp quality detector, a pulp quality detection method and a computer program product, and the pulp quality detector comprises a millimeter wave radar module which is used for detecting a to-be-detected fruit through millimeter waves; the processing module is connected with the millimeter-wave radar module so as to obtain echo signals obtained through detection from the millimeter-wave radar module, and three-dimensional reconstruction is carried out on the fruit to be detected based on the echo signals; wherein the three-dimensional fruit image after the three-dimensional reconstruction is used for carrying out quality detection on the pulp according to a preset analysis rule, and a pulp detection score is obtained.

Description

Technical Field

[0001] The present invention relates to the technical field of instrument detection, and in particular to a pulp quality detection instrument, a pulp quality detection method, and a computer program product. Background Art

[0002] The pulp quality testing of traditional thick-skinned fruits such as durian mainly relies on methods such as manual tapping and listening to sounds, and appearance observation. These methods are highly subjective, have low accuracy, and cannot intuitively judge the pulp content. Summary of the invention

[0003] The main purpose of the present invention is to provide a fruit pulp quality detector, a fruit pulp quality detection method, and a computer program product.

[0004] In order to achieve the above-mentioned purpose, according to a first aspect of the present invention, there is provided a fruit pulp quality detector comprising a millimeter wave radar module for detecting a fruit to be detected by means of millimeter waves; a processing module connected to the millimeter wave radar module to obtain an echo signal detected from the millimeter wave radar module, and to perform three-dimensional reconstruction of the fruit to be detected based on the echo signal; wherein the three-dimensional fruit image after three-dimensional reconstruction is used for quality detection of the fruit pulp according to preset analysis rules to obtain a fruit pulp detection score.

[0005] Optionally, it also includes a display module connected to the processing module, which is used to present an interactive page on the display screen of the display module, and the interactive page allows the user to control the opening and closing of the millimeter radar module in an interactive manner; the interactive page is also used to present the pulp quality detection results sent by the processing module.

[0006] Optionally, a memory is further included, wherein the memory stores a computer program that can be executed by the processing module, and the computer program is executed by the processing module to implement quality detection of the pulp based on preset analysis rules to obtain a pulp detection score.

[0007] Optionally, three-dimensional reconstruction of the fruit to be detected based on the echo signal includes: preprocessing the echo signal, wherein the preprocessing includes time domain denoising to remove random noise in the time domain signal; frequency domain filtering the signal after the time domain denoising to suppress unnecessary frequency components; spectral analysis of the preprocessed signal to separate reflection signals of different depths, wherein different millimeter waves penetrate the fruit skin and reflect different signals when encountering different media; and reconstructing a three-dimensional image of the inside of the fruit based on the separated reflection signals of different depths.

[0008] Optionally, performing quality inspection on the fruit pulp based on the three-dimensional fruit image after three-dimensional reconstruction and based on preset analysis rules includes: segmenting the three-dimensional fruit image based on a trained machine learning model to segment different structures of the fruit, wherein the different structures of the fruit include pulp, shell, cavity, and seeds; estimating the actual volume or weight of the segmented pulp area; and calculating the score of the pulp area based on preset quantitative indicators using a weighted summation method in combination with the actual volume or weight.

[0009] Optionally, estimating the actual volume of the segmented pulp area includes: determining the volume of the pulp area based on the number of voxels in the pulp area, or based on the surface area of ​​the pulp area determined by a triangular network constructed based on the pulp area; determining the average density of the pulp based on the type of fruit and the current maturity; determining the weight of the pulp based on the average density and the volume of the pulp area.

[0010] Optionally, based on a preset quantitative index, the score calculation of the pulp area is performed using a weighted summation method in combination with the actual volume or weight, including: determining the weight of the evaluation index for evaluating the pulp quality; based on the weight, volume or quality, the score calculation of the pulp area is performed using a weighted summation method to indicate the quality of the pulp.

[0011] According to a second aspect of the present invention, there is provided a method for detecting fruit pulp quality, comprising acquiring an echo signal of the detected fruit from a millimeter wave radar module; and performing three-dimensional reconstruction of the fruit to be detected based on the echo signal, wherein the three-dimensional fruit image after the three-dimensional reconstruction is used to perform quality detection on the fruit pulp according to preset analysis rules to obtain a fruit pulp detection score.

[0012] According to a third aspect of the present invention, there is provided a computer program product, comprising a computer program, which, when executed by a processing module, performs quality inspection on fruit pulp based on preset analysis rules to obtain a fruit pulp inspection score.

[0013] The present embodiment provides a fruit pulp quality detector, a fruit pulp quality detection method and a computer program product, wherein the fruit pulp quality detector comprises a millimeter wave radar module for detecting the fruit to be detected by means of millimeter waves; a processing module connected to the millimeter wave radar module to obtain the echo signal detected from the millimeter wave radar module, and to perform three-dimensional reconstruction of the fruit to be detected based on the echo signal; wherein the three-dimensional fruit image after three-dimensional reconstruction is used for quality detection of the fruit pulp according to preset analysis rules to obtain a fruit pulp detection score.

[0014] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0015] Figure 1 is a schematic diagram of a fruit pulp quality detector according to an embodiment of the present invention; DETAILED DESCRIPTION

[0016] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0017] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so as to describe the embodiments of the present invention described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0018] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0019] According to an embodiment of the present invention, a pulp quality detector is provided. Figure 1 As shown, it includes the following steps 101 to 103:

[0020] Step 101: A millimeter wave radar module is used to detect the fruit to be detected through millimeter waves.

[0021] In this step, a high-frequency millimeter-wave radar sensor may be used to transmit and receive millimeter-wave signals, penetrate the fruit shell or skin, and collect reflection data from the internal structure.

[0022] Step 102: A processing module is connected to the millimeter-wave radar module to obtain the echo signal detected by the millimeter-wave radar module, and perform three-dimensional reconstruction of the fruit to be detected based on the echo signal; wherein the three-dimensional fruit image after three-dimensional reconstruction is used for pre-set analysis rules to perform quality inspection on the fruit pulp and obtain a fruit pulp detection score.

[0023] In this step, the radar echo signal is received and converted into a 3D image of the inside of the fruit through signal processing algorithms and 3D reconstruction technology to show the distribution of the fruit pulp. Different from the echo of other objects, the echo of fruit has strong spectral characteristics due to its high moisture content, large scattering due to its irregular shape, and weak echo due to its softness. Based on these characteristics of the echo signal of the fruit, signal noise reduction and 3D reconstruction are performed.

[0024] As an optional implementation of this embodiment, a power supply module is also included to provide power supply.

[0025] As an optional implementation method of this embodiment, the fruit pulp quality detector also includes a display module, which presents an interactive page on the display screen of the display module. The interactive page allows the user to control the opening and closing of the millimeter radar module in an interactive manner; the interactive page is also used to present the fruit pulp quality detection results.

[0026] In this optional implementation, the device uses a touch screen as the user interface. The interface design is simple and easy to understand. The main interface is equipped with function buttons such as "Start Detection", "View Results" and "Settings". The user only needs to bring the device close to the surface of the fruit and click "Start Detection". The device will automatically scan, analyze and display the pulp percentage and quality grade. In order to improve the user experience, a voice broadcast function has been added, and the test results can be reported verbally after the test is completed.

[0027] In this optional implementation, the millimeter wave radar, processing module, power module and display screen are integrated into a handheld device to obtain a portable fruit pulp detector, which is convenient to carry and operate with one hand.

[0028] As an optional implementation method of this embodiment, three-dimensional reconstruction of the fruit to be detected based on the echo signal includes: preprocessing the echo signal, wherein the preprocessing includes time domain denoising to remove random noise in the time domain signal; frequency domain filtering the signal after the time domain denoising to suppress unnecessary frequency components; spectral analysis of the preprocessed signal to separate reflection signals of different depths, wherein different millimeter waves penetrate the fruit skin and reflect different signals when encountering different media; and reconstructing a three-dimensional image of the inside of the fruit based on the separated reflection signals of different depths.

[0029] In this optional implementation, noise reduction includes time domain denoising: applying a Gaussian filter or a median filter to remove random noise in the time domain signal; frequency domain filtering: using Fourier transform to convert the signal to the frequency domain, and suppressing unnecessary frequency components through a bandpass filter (Butterworth or Chebyshev filter), thereby reducing the impact of stray signals.

[0030] When performing three-dimensional reconstruction, based on the tomographic imaging method, multi-angle or multi-frequency data is used for three-dimensional reconstruction, and then a point cloud representing each layer of the fruit is generated. After that, surface reconstruction is performed, and a surface mesh model is constructed using the point cloud data to form a complete three-dimensional image of the fruit.

[0031] Exemplarily, the millimeter wave radar module uses a continuous wave radar sensor with an operating frequency of 24GHz, which has the characteristics of strong penetration ability and high resolution, and is suitable for the detection of fruits such as durian that have a hard shell and contain moisture inside. The radar sensor sends millimeter wave pulses through the antenna array. After these pulses penetrate the durian shell, they encounter different media (such as meat, cavity) and will be reflected. The reflected signal is received by the same antenna array. The signal processing unit first pre-processes the received radar echo signal, including amplification, filtering and analog-to-digital conversion (ADC) to ensure signal quality. Subsequently, the fast Fourier transform (FFT) is used to perform spectrum analysis on the signal to separate the reflected signals at different depths.

[0032] As an optional implementation method of this embodiment, quality inspection of the fruit pulp based on the three-dimensional fruit image after three-dimensional reconstruction based on preset analysis rules includes: segmenting the three-dimensional fruit image based on a trained machine learning model to segment different structures of the fruit; estimating the actual volume or weight of the segmented pulp area; and calculating the score of the pulp area based on preset quantitative indicators using a weighted summation method combined with the actual volume or weight.

[0033] In this optional implementation, the pulp area in the image is automatically identified, the pulp volume or weight ratio is calculated, and the pulp quality evaluation is output in combination with the preset quality standard.

[0034] When automatically identifying the pulp area, the Otsu threshold method or other adaptive threshold techniques can be used to automatically determine the boundary between the pulp and non-pulp areas; feature extraction: calculate image texture features (such as gray-level co-occurrence matrix GLCM, local binary pattern LBP), shape features (such as circularity, aspect ratio) and statistical features (such as mean, standard deviation) to distinguish the pulp from other tissues; machine learning classification: apply support vector machine (SVM), random forest or neural network to classify the extracted features to accurately distinguish the pulp from other parts such as seeds and peel.

[0035] A machine learning algorithm, specifically a convolutional neural network (CNN), is used to automatically segment the flesh area in the image. The training dataset consists of radar imaging samples of various fruits with known flesh content and quality. The model is trained through deep learning to enable it to accurately distinguish between flesh, shell and cavity.

[0036] As an optional implementation method of the present embodiment, estimating the actual volume of the segmented pulp area includes: determining the volume of the pulp area based on the number of voxels in the pulp area, or the surface area of ​​the pulp area determined based on a triangular network constructed based on the pulp area; determining the average density of the pulp based on the type of fruit and the current maturity; determining the weight of the pulp based on the average density and the volume of the pulp area.

[0037] In this optional implementation, based on the segmented pulp area, the three-dimensional image data is used to calculate the actual volume of the pulp or estimate its weight. The specific method for volume calculation is to divide the three-dimensional image into multiple voxels, calculate the number of voxels in the pulp area, each voxel represents a certain volume, and accumulate them to obtain the volume; or construct the surface of the pulp area into a triangular mesh, and then calculate the area of ​​each triangle, and accumulate them to obtain the total area of ​​the pulp surface, and then calculate the approximate volume based on the surface mesh; or use a numerical integration method (Monte Carlo integration or Simpson's rule) to calculate the volume of the pulp area.

[0038] The weight algorithm is to pre-determine the average density of the pulp according to the type and maturity of the pulp, and multiply the calculated pulp volume by the average density of the pulp to obtain the weight of the pulp. Quality assessment also includes quantification of quality indicators: factors such as pulp uniformity, density, and moisture content are converted into quantitative indicators according to industry standards or user-defined standards; comprehensive scoring: the above indicators are weighted and summarized to form the final pulp quality score.

[0039] As an optional implementation method of this embodiment, using a weighted summation method in combination with the actual volume or weight to calculate the score of the pulp area includes: determining the weight of the evaluation index used to evaluate the pulp quality; based on the weight, volume or mass, using a weighted summation method to calculate the score of the pulp area to indicate the quality of the pulp.

[0040] In this optional implementation, during the recognition process, the algorithm classifies each pixel point and eventually forms the precise boundary of the pulp area. In this process, the empty house detection of durian can be completed. In terms of quality analysis, the overall pulp volume or weight ratio is calculated based on the segmented pulp area, and the evaluation index is set to grade the quality of durian in combination with the preset quality judgment criteria. For example, the maturity and taste of durian are evaluated by analyzing the uniform distribution of the pulp area and the average reflection intensity (reflecting the water content), and the quality of durian is determined based on the empty house detection results. These calculation and analysis results are displayed to the user in intuitive graphics and text through the user interface.

[0041] The millimeter wave radar technology in this embodiment can accurately capture the fine structure inside the fruit, significantly improving the detection accuracy compared to traditional methods. It avoids the loss caused by fruit opening detection and protects the value of the product. It can achieve instant detection and obtain results, shorten the detection time, and improve circulation efficiency. It is applicable to various varieties of fruits and meets the needs of different user groups.

[0042] According to an embodiment of the present invention, a method for detecting fruit pulp quality is also provided, comprising acquiring an echo signal of the detected fruit from a millimeter wave radar module; and performing three-dimensional reconstruction of the fruit to be detected based on the echo signal, wherein the three-dimensional fruit image after the three-dimensional reconstruction is used to perform quality detection on the fruit pulp according to preset analysis rules to obtain a fruit pulp detection score.

[0043] According to an embodiment of the present invention, the present invention also provides an electronic device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can implement the method described in any of the above embodiments when executing.

[0044] According to an embodiment of the present invention, the present invention further provides a readable storage medium, which stores computer instructions, and the computer instructions are used to enable a computer to implement the method described in any of the above embodiments when executed.

[0045] According to an embodiment of the present invention, the present invention further provides a computer program product, which can implement the method described in any of the above embodiments when executed by a processor.

[0046] The program code for implementing the method of the present invention may be written in any combination of one or more programming languages. These program codes may be provided to a processing module of a pulp quality detection instrument so that when the program code is executed by a processor or controller, the functions / operations specified by the method are implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

Claims

1. A pulp quality detector, characterized in that: include: Millimeter wave radar module, used to detect the fruit to be inspected through millimeter waves; A processing module connected to the millimeter wave radar module to obtain the echo signal detected by the millimeter wave radar module, and perform three-dimensional reconstruction of the fruit to be detected based on the echo signal; The three-dimensional fruit image after three-dimensional reconstruction is used to perform quality inspection on the fruit pulp according to preset analysis rules to obtain a fruit pulp inspection score.

2. The pulp quality detector according to claim 1, characterized in that: Also included is a display module connected to the processing module, which is used to present an interactive page on a display screen of the display module, wherein the interactive page allows a user to control the opening and closing of the millimeter radar module in an interactive manner; The interactive page is also used to present the pulp quality detection results sent by the processing module.

3. The pulp quality detector according to claim 1, characterized in that: It also includes a memory, which stores a computer program that can be executed by the processing module. The computer program is executed by the processing module to implement quality detection of the pulp based on preset analysis rules to obtain a pulp detection score.

4. The pulp quality detector according to claim 1, characterized in that: Performing three-dimensional reconstruction of the fruit to be detected based on the echo signal includes: Preprocessing the echo signal, wherein the preprocessing includes time domain denoising to remove random noise in the time domain signal; performing frequency domain filtering on the signal after the time domain denoising to suppress unnecessary frequency components; Perform spectrum analysis on the pre-processed signal to separate the reflected signals at different depths. After penetrating the fruit skin, different millimeter waves encounter different media and reflect different signals. The three-dimensional image of the interior of the fruit is reconstructed based on the separated reflection signals at different depths.

5. The pulp quality detector according to claim 1, characterized in that: The quality detection of the fruit pulp based on the three-dimensional fruit image after three-dimensional reconstruction and the preset analysis rules includes: Segmenting the three-dimensional fruit image based on the trained machine learning model to segment different structures of the fruit, wherein the different structures of the fruit include flesh, shell, cavity, and seeds; Estimate the actual volume or weight of the segmented pulp area; Based on the preset quantitative index, the score of the pulp area is calculated by using a weighted summation method in combination with the actual volume or weight.

6. The pulp quality detector according to claim 3, characterized in that: Estimating the actual volume of the segmented pulp area includes: Determine the volume of the flesh region based on the number of voxels in the flesh region or the surface area of ​​the flesh region determined based on a triangulated network constructed based on the flesh region; Determine the average density of the flesh based on the type of fruit and current maturity; The weight of the pulp is determined based on the average density and the volume of the pulp area.

7. The pulp quality detector according to claim 4, characterized in that: Based on the preset quantitative index, the score calculation of the pulp area is performed by combining the actual volume or weight using a weighted summation method, including: Determine the weights of the evaluation criteria used to evaluate pulp quality; Based on the weight, volume or mass, a weight summation method is used to calculate the score of the pulp area to indicate the quality of the pulp.

8. The pulp quality detector according to claim 1, characterized in that: It also includes a power module for providing power supply.

9. A method for detecting pulp quality, characterized in that: include: Acquire the echo signal of the detected fruit from the millimeter wave radar module; The fruit to be inspected is three-dimensionally reconstructed based on the echo signal, wherein the three-dimensional fruit image after the three-dimensional reconstruction is used to perform quality inspection on the fruit pulp according to a preset analysis rule to obtain a fruit pulp inspection score.

10. A computer program product, characterized in that It includes a computer program, which, when executed by a processing module, performs quality inspection on the fruit pulp based on preset analysis rules to obtain a fruit pulp inspection score.

Images