Economic fruit automatic quality evaluation model construction system, modeling method and application thereof

Abstract

The invention relates to an economic forest fruit automatic quality evaluation model construction system, which is provided with an image spectrum acquisition area configured to nondestructively acquire image information and spectrum information of fruits; the hardness information acquisition area is configured to acquire hardness information of the fruits; the sugar degree information acquisition area is configured to acquire sugar degree information of fruits; the device further comprises a conveying mechanism and computer equipment, the conveying mechanism is provided with a flexible fruit support used for bearing fruits, and the image spectrum collecting area, the hardness information collecting area and the sugar degree information collecting area are sequentially arranged in the conveying direction of the conveying mechanism. The conveying mechanism conveys the fruits to the image spectrum collection area, the hardness information collection area and the sugar degree information collection area according to a preset instruction to sequentially collect image information, spectrum information, hardness information and sugar degree information; and the computer equipment constructs an automatic fruit quality evaluation model according to the acquired image information, spectral information, hardness information and sugar degree information. According to the system, the whole process of internal and external quality information collection and prediction model construction of the persimmon fruits can be automatically completed, manpower participation in the fruit quality modeling process is reduced, and the intelligent degree of grading and sorting of the persimmon fruits is improved.

Description

Technical Field

[0001] The present application belongs to the field of artificial intelligence model modeling technology, and in particular relates to a fully automated internal and external quality modeling system for fruits, and specifically relates to an automated quality evaluation model construction system for economic fruit trees, and its automated modeling method and application. Background Art

[0002] The construction of quality prediction models is a crucial step in the automated nondestructive testing and grading of economic fruit trees. Traditional fruit quality modeling relies heavily on manual labor, resulting in low efficiency and high labor costs. While some automated fruit quality modeling technologies have emerged with technological advancements, there is still a lack of effective devices and methods that can automate the entire process of collecting internal and external quality data and building prediction models for fruit. Existing testing methods struggle to simultaneously and efficiently capture multiple aspects of fruit information, including images, spectra, firmness, and sugar content, and are therefore unable to adequately meet the modern fruit industry's demand for automated quality modeling.

[0003] Therefore, there is an urgent need to design a device and method for automated internal and external quality modeling of economic fruit trees to solve the above technical problems, and the present invention is therefore provided. Summary of the Invention

[0004] The embodiments of the present application provide a modeling system and method for an automated internal and external quality evaluation model of fruit to address the problems that existing fruit quality modeling requires a large amount of manpower, has low modeling efficiency, high labor costs, and cannot well adapt to the modern fruit industry's demand for automated quality modeling.

[0005] In a first aspect, an embodiment of the present application provides a system for constructing an automated quality evaluation model for economic fruit trees, which is provided with:

[0006] An image spectrum acquisition area is configured to non-destructively acquire image information and spectral information of fruits;

[0007] a hardness information collection area configured to collect hardness information of fruits;

[0008] The sugar content information collection area is configured to collect sugar content information of fruits;

[0009] The device also includes a conveying mechanism and a computer device. The conveying mechanism is provided with a flexible fruit tray for carrying fruit. The image spectrum collection area, the hardness information collection area, and the sugar content information collection area are sequentially arranged along the conveying direction of the conveying mechanism. The conveying mechanism conveys the fruit to the image spectrum collection area, the hardness information collection area, and the sugar content information collection area according to preset instructions to collect image information and spectrum information, hardness information, and sugar content information in sequence.

[0010] The computer device constructs an automatic fruit quality evaluation model based on the collected image information, spectral information, hardness information and sugar content information.

[0011] In a second aspect, an embodiment of the present application provides a method for modeling a fully automated fruit quality evaluation model using the system of the present invention, characterized in that it includes the following steps:

[0012] (S1) placing the fruit to be inspected on a flexible fruit tray, using a conveying mechanism to convey the fruit to an image spectrum acquisition area, and using an image acquisition mechanism and a spectrum acquisition mechanism in a dark box to omnidirectionally and losslessly collect image information and spectrum information of the fruit;

[0013] (S2) using a conveying mechanism to sequentially convey the fruit to a hardness information collection area and a sugar content information collection area, wherein the hardness information of the fruit is collected by the hardness collection mechanism, and the sugar content information of the fruit is collected by the sugar content collection mechanism;

[0014] (S3) Constructing an internal and external quality prediction model of the fruit to be tested based on the image information, spectral information, hardness information and sugar content information of the fruit.

[0015] On the third aspect, the embodiment of the present application provides an application of an economic fruit and vegetable automated quality evaluation model construction system in modeling a fully automated fruit quality evaluation model.

[0016] The economic fruit automated quality evaluation model construction system of the present invention can automatically complete the entire process of collecting internal and external quality information of fruits and constructing prediction models, reduce human participation in the fruit quality modeling process, and improve the intelligence level of fruit grading and sorting. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0018] Figure 1 A schematic diagram of the three-dimensional structure of the system for constructing the automated quality evaluation model of economic fruit trees in this application;

[0019] Figure 2 Another structural diagram of the system for constructing an automated quality evaluation model for economic fruit trees in this application;

[0020] Figure 3 This is another three-dimensional structural diagram of the system for constructing an automated quality evaluation model for economic fruit trees in this application;

[0021] Figure 4 A schematic diagram of the structure of the transmission mechanism provided in an embodiment of the present application;

[0022] Figure 5 A schematic diagram of the specific installation structure of the transmission mechanism provided in an embodiment of the present application;

[0023] Figure 6 A schematic diagram of the structure inside the dark box provided in an embodiment of the present application;

[0024] Figure 7 A schematic diagram of the structure of the image acquisition mechanism in the darkroom provided in an embodiment of the present application;

[0025] Figure 8 A schematic diagram of the structure of the fill light mechanism in the dark box provided in an embodiment of the present application;

[0026] Figure 9 Another structural diagram of the fill light mechanism in the dark box provided by an embodiment of the present application;

[0027] Figure 10 This is another structural diagram of the fill light mechanism in the dark box provided by an embodiment of the present application;

[0028] Figure 11 A schematic cross-sectional view of a halogen light source in a fill light mechanism in a darkroom according to an embodiment of the present application;

[0029] Figure 12 A schematic diagram of the structure of a fruit turning mechanism in a dark box provided in an embodiment of the present application;

[0030] Figure 13 This is another structural diagram of the fruit turning mechanism in the dark box provided by an embodiment of the present application;

[0031] Figure 14 A schematic diagram of the structure of the hardness collection mechanism provided in an embodiment of the present application;

[0032] Figure 15 A schematic diagram of the three-dimensional structure of the hardness collection mechanism provided in an embodiment of the present application;

[0033] Figure 16 This is another schematic diagram of the three-dimensional structure of the hardness collection mechanism provided in an embodiment of the present application;

[0034] Figure 17 A schematic diagram of the three-dimensional structure of the automatic juice extraction mechanism and the refractometer measurement mechanism provided in an embodiment of the present application;

[0035] Figure 18 A schematic diagram of the structure of a juice extractor provided in an embodiment of the present application;

[0036] Figure 19 A schematic diagram of the three-dimensional structure of a refractometer provided in an embodiment of the present application;

[0037] Figure 20 A block diagram of the computer device provided in accordance with an embodiment of the present application;

[0038] Figure 21 This is a spectrum information diagram of a persimmon sample in an embodiment of the present invention;

[0039] Figure 22 This is a hardness information diagram of a persimmon sample in an embodiment of the present invention;

[0040] Figure 23 This is a sugar content information diagram of a persimmon sample in an embodiment of the present invention;

[0041] Figure 24 This is the image information and recognition result diagram of the persimmon sample in the embodiment of the present invention;

[0042] Figure 25 This is a graph showing the prediction results of the prediction model established in an embodiment of the present invention for the hardness of persimmon samples;

[0043] Figure 26 This is a graph showing the prediction results of the prediction model established in an embodiment of the present invention for the sugar content of persimmon samples. DETAILED DESCRIPTION

[0044] The features and exemplary embodiments of various aspects of the present application will be described in detail below. In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present application, rather than to limit the present application. For those skilled in the art, the present application can be implemented without the need for some of these specific details. The following description of the embodiments is merely to provide a better understanding of the present application by illustrating the examples of the present application.

[0045] In the existing technology, since fruit quality modeling relies on temporary manual quality control, it is impossible to effectively carry out fully automated quality control. Figures 1 to 3As shown, an embodiment of the present invention provides a system for constructing an automated quality evaluation model for economic fruit trees. The system comprises: an image spectrum acquisition area 1 configured to non-destructively acquire image and spectral information of fruit; a hardness information acquisition area 2 configured to acquire hardness information of the fruit; and a sugar content information acquisition area 3 configured to acquire sugar content information of the fruit. The system also includes a conveyor mechanism 4 and a computer device (not shown). The conveyor mechanism 4 is provided with a flexible fruit tray 5 for carrying the fruit. The image spectrum acquisition area 1, the hardness information acquisition area 2, and the sugar content information acquisition area 3 are sequentially arranged along the conveying direction of the conveyor mechanism 4. The conveyor mechanism 4, according to preset instructions, conveys the fruit to the image spectrum acquisition area 1, the hardness information acquisition area 2, and the sugar content information acquisition area 3, where they sequentially acquire image and spectral information, as well as hardness and sugar content information. The computer device constructs an automated fruit quality evaluation model based on the acquired image and spectral information, hardness information, and sugar content information. The system can automatically complete the acquisition of internal and external fruit quality and the construction of a prediction model, enabling non-destructive prediction of internal and external fruit quality, significantly reducing manual intervention. The present invention automatically transports the flexible fruit tray 5 through the conveying mechanism 4. The fruit passes through the image spectrum collection area 1, the hardness information collection area 2, and the sugar content information collection area 3 in sequence. Under the action of the corresponding sensors and controllers in each area, the fruit operates in an orderly manner, realizing the rapid collection and modeling of a large amount of data on the fruit, and providing technical support for the subsequent intelligent grading and sorting of the fruit.

[0046] It should be noted that the height direction of the transmission mechanism is the first direction DR1, the width direction of the transmission mechanism is the second direction DR2, and the length direction of the transmission mechanism is the third direction DR3. The third direction DR3 is the normal direction of the plane defined by the first direction DR1 and the second direction DR2, that is, the length direction or transmission direction of the transmission mechanism. The directions indicated by the first, second and third directions DR1, DR2, and DR3 described in this specification are relative concepts and can be changed to other directions.

[0047] like Figures 1 to 3As shown, image spectrum acquisition area 1 is provided with an image acquisition mechanism 11 and a spectrum acquisition mechanism 13 for omnidirectional, lossless acquisition of fruit images and spectral information. Hardness information acquisition area 2 is provided with a hardness acquisition mechanism 21 for acquiring fruit hardness information. Sugar content acquisition area 3 is provided with a sugar content acquisition mechanism 31 for acquiring fruit sugar content information. In the specific technical solution of the present invention, persimmons are used as an example of economic fruit. Berries and pome fruits can also be constructed and applied using the automated quality evaluation model construction system for economic fruit according to the present invention. Specific economic fruits include thin-skinned, seedless fruits such as persimmons, kiwis, and blueberries, as well as thin-skinned fruits with small cores that can be directly juiced, such as apples and pears. Persimmons are used as a typical example for illustration. A conveyor mechanism 4 horizontally transports a flexible fruit tray 5. An image spectrum acquisition area 1, a hardness information acquisition area 2, and a sugar content information acquisition area 3 are sequentially arranged along the conveying direction of the conveyor mechanism 4. The flexible fruit tray 5 is placed on the conveyor mechanism 4, and the fruit to be tested is placed on the flexible fruit tray 5. In this way, the conveyor mechanism 4 can sequentially transport persimmons and other fruits between the image spectrum acquisition area 1, the hardness information acquisition area 2, and the sugar content information acquisition area 3. The system combines image information, spectral information, hardness information, and sugar content information, and uses image / spectral analysis methods to construct an internal and external quality prediction model for the fruit to be tested based on the image information, spectral information, hardness information, and sugar content information of the fruit. The present invention collects and evaluates the internal and external quality of fruit products from multiple dimensions, provides an intelligent modeling scenario for the fruit quality assessment model, and provides strong support for subsequent fruit quality optimization and intelligent sorting.

[0048] The following is a detailed description of each component in the system for building an automated quality evaluation model for economic fruit trees.

[0049] <Transmission mechanism 4 part>

[0050] like Figure 4 and Figure 5 As shown, the conveying mechanism 4 includes a main frame 41, a conveyor belt assembly 43 arranged in the main frame 41, and a conveying motor 42 fixedly installed with the main frame 41. The flexible fruit tray 5 is installed on the conveyor belt assembly 43 and is moved by the conveyor belt assembly 43 to the image spectrum collection area 1, the hardness information collection area 2, and the sugar content information collection area 3 according to preset instructions.

[0051] The conveyor belt assembly 43 includes two parallel conveyor belts 433 and several conveyor guide wheels arranged in sequence along the conveying direction of the conveying mechanism 4; the conveyor guide wheels are connected to the main frame 41 of the conveying mechanism, including a first conveyor wheel 431 and a second conveyor wheel 432 coaxially connected, one of the two conveyor belts 433 is sequentially sleeved on each first conveyor wheel 431, and the other conveyor belt 433 is sequentially sleeved on each second conveyor wheel 432, and the two conveyor belts 433 are driven by the conveying motor 42 to move synchronously according to preset instructions; the output shaft of the conveying motor 42 is connected to one of the several conveyor guide wheels, and the two ends of the flexible fruit tray 5 are respectively connected to the two conveyor belts 433.

[0052] like Figure 5 As shown, the conveyor guide wheel in the conveyor mechanism 4 has a conveyor guide wheel shaft 434. The first conveyor wheel 431 and the second conveyor wheel 432 are mounted on the same conveyor guide wheel shaft 434. Each conveyor guide wheel shaft 434 is provided with a conveyor guide wheel shaft support 435 at both ends. Several pairs of conveyor guide wheel shaft supports 435 are evenly spaced along the extension direction of the conveyor mechanism 4 on the main frame 41. Each pair of conveyor guide wheel shaft supports 435 is symmetrically arranged on both sides of the main frame 41 along the width direction of the conveyor mechanism 4. A conveyor guide wheel shaft 434 is disposed between each pair of conveyor guide wheel shaft supports 435. One of the conveyor guide wheel shafts 434 is connected to the output shaft of the conveyor motor 42, which transmits the operating power of the conveyor belt 433. Both ends of each conveying guide wheel shaft 434 are provided with conveying guide wheels. The conveying guide wheels on the same side along the length direction of the main frame 41 are connected by a conveyor belt 433. The flexible fruit tray 5 is arranged between the two conveyor belts 433 or fixed to the two conveyor belts 433, and moves with the movement of the conveyor belt 433.

[0053] <Image spectrum acquisition area 1>

[0054] like Figure 3 、 Figures 6 to 13 As shown, the image spectrum acquisition area 1 is provided with a dark box 14, within which are disposed an image acquisition mechanism 11, a spectrum acquisition mechanism 13, and a fill light mechanism 12. The fill light mechanism 12 is disposed on a side of the flexible fruit holder 5 close to the image acquisition mechanism 11, while the spectrum acquisition mechanism 13 is disposed on a side of the flexible fruit holder 5 away from the image acquisition mechanism 11. The present invention utilizes the dark box 14 to non-destructively acquire image and spectrum information of the fruit in all directions, thereby preventing interference from ambient light during the acquisition of the image and spectrum information.

[0055] The dark box 14 includes a dark box body and a driving mechanism arranged in the dark box body for driving the image acquisition mechanism 11 and the fill light mechanism 12 to slide. The driving mechanism includes a first screw rod 141 and a second screw rod 142 installed in the dark box body. Specifically, the dark box body includes a first frame 140 and a plurality of light shielding plates 143; the light shielding plates 143 are fixedly connected to the outer periphery of the first frame 140 and are used for shading inside the dark box body. The first screw rod 141 and the second screw rod 142 are rotatably connected to the two sides of the first frame 140 respectively. The first screw rod 141 and the second screw rod 142 are respectively connected to two screw motors. The screw motors can drive the first screw rod 141 and the second screw rod 142 to rotate around their own axes, thereby driving the image acquisition mechanism 11 and the fill light mechanism 12 to move in the first direction.

[0056] The image acquisition mechanism 11 and the spectrum acquisition mechanism 13 are arranged relative to each other in a first direction, and the fill light mechanism 12 is located between the image acquisition mechanism 11 and the spectrum acquisition mechanism 13. The image capture path of the image acquisition mechanism 11, the light source supplement path of the fill light mechanism 12, and the spectrum acquisition path of the spectrum acquisition mechanism 13 overlap. When performing image and spectrum acquisition, the image acquisition mechanism 11, the fill light mechanism 12, the spectrum acquisition mechanism 13, and the flexible fruit tray 5 are coaxially arranged in the first direction.

[0057] The image acquisition mechanism 11 includes a distance sensor 111 for sensing the distance of the fruit on the flexible fruit tray, an industrial camera 112 for imaging the fruit to be inspected, and a camera slide 113 for adjusting the relative position of the industrial camera 112. The industrial camera 112 can move in a third direction. The camera slide 113 is connected to a second screw rod 142 for vertical movement. The industrial camera 112 and the distance sensor 111 are both located on the side of the camera slide 113 relative to the flexible fruit tray 5. The industrial camera 112 and the distance sensor 111 are both located relative to the flexible fruit tray 5. The distance sensor 111 is used to monitor the distance between the industrial camera 112 and the fruit, thereby better imaging and focusing the fruit. The industrial camera 112 is used to image the fruit to be inspected.

[0058] A typical example of the spectrum acquisition mechanism 13 of the present invention is a spectrometer. The position of the spectrum acquisition mechanism 13 is relatively fixed. For example, it can be set at a fixed position in the image spectrum acquisition area 1. When the flexible fruit tray moves to the image spectrum acquisition area 1, the acquisition probe of the spectrometer is set directly facing the fruit. Typically, the spectrometer is fixedly mounted on the main frame 41 of the conveying mechanism 4. In this way, in the darkroom 14, the image acquisition mechanism 11, the fill light mechanism 12 and the spectrum acquisition mechanism 13 are coaxially arranged from top to bottom. When the flexible fruit tray 5 moves to the preset position of the darkroom 14, the flexible fruit tray 5 is located above the spectrometer, thereby collecting the spectral information of the fruit.

[0059] As Figure 8-10As shown, the fill light mechanism 12 includes a movable fill light platform 121, an outer light shield 122 fixed to the fill light platform 121, and an inner fill light ring 123 embedded in the inner side of the outer light shield 122; the outer light shield 122 and the inner fill light ring 123 are coaxially arranged, and a first pulley motor 124 is provided on the outer light shield 122. The output shaft of the first pulley motor 124 is connected to the inner ring transmission belt 126 sleeved on the inner fill light ring 123 through a first small pulley 125, driving the inner fill light ring 123 to rotate and switch between the camera fill light position and the spectrum fill light position.

[0060] Several halogen light sources 129 are evenly distributed circumferentially on the side of the inner fill light ring 123 facing the transmission mechanism. The outer light shield 122 includes an integrally formed inner and outer rings. The halogen light sources 129 are positioned between the inner and outer rings of the outer light shield 122. The inner ring of the outer light shield 122 is provided with several U-shaped holes 127 for spectral fill light, while the outer ring of the outer light shield 122 is provided with several U-shaped holes 128 for camera fill light. The U-shaped holes 127 for spectral fill light are interlaced with the U-shaped holes 128 for camera fill light. In a preferred embodiment, an annular guide rail is provided between the outer light shield 122 and the inner fill light ring 123. The halogen light sources 129 are arranged around the annular guide rail and can switch between two different operating modes: spectral fill light and camera fill light by rotating with the first pulley motor 124. The use of an annular guide rail between the inner fill light ring 123 and the outer light shield 122 for sliding facilitates the switching of the halogen light source 129 between the spectral fill light and camera fill light working modes, and effectively utilizes the same light source to meet the needs of image acquisition and spectral acquisition, avoiding the mutual interference between the two types of information acquisition. The camera fill light U-shaped hole 128 formed by the outer ring groove can create a soft fill light environment, avoiding the presence of a large number of overexposed images in the image information caused by the halogen light source 129 directly projected onto the surface of the fruit during image acquisition. The spectral fill light U-shaped hole 127 formed by the inner ring groove avoids the problem of the halogen light source 129 interfering with the spectrometer's collection of the internal quality information of the fruit through the diffuse reflection plane of the inner surface of the light shielding structure of the dark box during spectral information acquisition.

[0061] The inner fill light ring 123 may be annular, and the orthographic projections of the industrial camera 112 and the distance sensor 111 on the transmission mechanism are located inside the orthographic projections of the inner fill light ring 123 on the transmission mechanism, facilitating the industrial camera 112 to collect image information.

[0062] The filler slide 121 is connected to a first screw 141 for vertical movement. The filler slide 121 is positioned below the camera slide 113. An inner filler light ring 123 is coaxially mounted on the filler slide 121 via an annular guide rail, allowing for rotation about its own axis. The axis of the inner filler light ring 123 is parallel to the vertical direction. Several pairs of halogen light sources 129 are evenly spaced along the circumference of the inner filler light ring 123. A first small pulley 125 is connected to the outer circumference of the inner filler light ring 123 via an inner ring transmission belt 126. The first small pulley 125 is externally connected to a first pulley motor 124, which drives the first small pulley 125 to rotate. This in turn, through friction from the inner ring transmission belt 126, drives the inner filler light ring 123 to rotate about its own axis. The present invention utilizes the wide color temperature range of the halogen light source 129 to provide effective light source support for asynchronous acquisition of spectral and image information, improving light source utilization efficiency and simplifying the overall structural design within the darkroom 14.

[0063] like Figure 8 and Figure 9 As shown, a circle of camera fill light U-shaped holes 128 is provided on the outer side of the bottom of the outer light shield 122, and a circle of spectrum fill light U-shaped holes 127 is provided on the inner side of the bottom. The camera fill light U-shaped holes 128 and the spectrum fill light U-shaped holes 127 are arranged alternately.

[0064] like Figure 11 As shown, the halogen light source 129 includes a halogen light source housing, a lamp head 116, a focus cover 117 and a beam splitter 118. The lamp head 116 and the focus cover 117 are coaxially connected to the inner side of the halogen light source housing. The beam splitter 118 is connected to the inner bottom surface of the halogen light source housing. The beam splitter 118 is used to split the light beam emitted by the halogen light source 129 into two beams of light that are respectively irradiated toward the outer circle and the inner circle of the outer light shield 122; according to the work station of the halogen light source 129, the light separated by the beam splitter 118 is filled into persimmon fruits by the camera fill light U-shaped hole 128 or the spectrum fill light U-shaped hole 127.

[0065] Lamp head 116 is located inside condenser 117. The inner bottom surface of the halogen light source housing is transparent. A beam splitter 118 is used to split the light beam emitted by halogen light source 129 into two beams. Depending on the position of halogen light source 129, these two beams of light illuminate either camera fill light U-shaped aperture 128 or spectrum fill light U-shaped aperture 127. When halogen light source 129 is in the camera fill light position, beam splitter 118 is used to project the light beam emitted by halogen light source 129 through camera fill light U-shaped aperture 128 onto the inner surface of light shielding plate 143 within the dark box. The surface of light shielding plate 143 generates diffuse reflection, and the diffusely reflected light is then projected onto the surface of the fruit. This creates a softer and more uniform fill light on the fruit surface, facilitating image acquisition by industrial cameras. When the halogen light source 129 is in the spectral fill light position, the spectrometer 118 is used to project the light beam emitted by the halogen light source 129 directly onto the surface of the fruit through the spectral fill light U-shaped hole 127 to avoid the influence of diffuse reflection from the light shielding plate in the dark box.

[0066] In a preferred embodiment of the present invention, the image spectrum acquisition area 1 is further provided with a fruit turning mechanism 15 , which is configured to turn over the fruit placed on the flexible fruit tray 5 to expose different surfaces of the fruit.

[0067] like Figure 3 and Figure 5 As shown, combined with Figure 12 and Figure 13 As shown, there are two fruit turning mechanisms 15 arranged in the image spectrum acquisition area 1 , and the two fruit turning mechanisms 15 are arranged opposite to each other in the second direction.

[0068] The fruit turning mechanism 15 includes a first contact plate 151 and a first pushing mechanism 150 fixed on the transmission mechanism. The first contact plate 151 is installed at one end of the first pushing mechanism 150 and is configured to contact the persimmon fruit and limit the fruit; the first pushing mechanism 150 is configured to push the persimmon fruit on the flexible fruit tray 5 to replace the exposed surface of the persimmon fruit.

[0069] The image spectrum collection area 1 is further provided with a first sensing mechanism 16, which is configured to sense whether the flexible fruit tray 5 has entered a specific position within the image spectrum collection area 1. Specifically, the first sensing mechanism 16 comprises two opposing photoelectric sensors fixedly mounted on the main frame 41 of the transmission mechanism. The two photoelectric sensors are used to monitor the rest position of the flexible fruit tray 5 within the image spectrum collection area 1.

[0070] The first pushing mechanism 150 includes a first flipping bracket 152 and a first rotating motor 153 . The first rotating motor 153 is installed on the first flipping bracket 152 and is configured to drive the first contact plate 151 to rotate so as to flip the fruit.

[0071] A first transmission rack 154 is provided in the first flip bracket 152, and the lower ends of the first transmission rack 154 are respectively engaged with a first driving gear 155 and a first driven gear 156 of the same outer diameter, one end of which is fixed to the first contact plate 151. The first driving gear 155 and the first driven gear 156 are installed in the first flip bracket 152, and the first driving gear 155 is connected to the output shaft of the first worm gear motor 157, driving the first transmission rack 154 to move in the second direction.

[0072] like Figure 12 and Figure 13 As shown, the first pushing mechanism 150 also includes a first push rod base 158 fixed to the transmission mechanism main frame 41. This first push rod base 158 is connected to the first flip bracket 152 and can push the first flip bracket 152 to move in a first direction, thereby adjusting the height of the first pushing mechanism 150 according to the specific height of the fruit. The first sensing mechanism 16 is mounted on the first push rod base 158. The first push rod bases 158 of the two fruit flipping mechanisms 15 are respectively located on either side of the first frame 140 and below the fill light mechanism 12.

[0073] The image acquisition mechanism 11 and the fill light mechanism 12 are both connected to the first frame 140. The lower end of the first flipping bracket 152 is fixed to the main frame 41 of the conveyor mechanism 4 via a first push rod base 158. The two fruit flipping mechanisms 15 are symmetrically arranged along the horizontal central axis of the dark box 14 to facilitate contact flipping of the fruit.

[0074] The camera slide 113 and the fill-in light slide 121 are connected to a drive mechanism to drive the image acquisition mechanism 11 and the fill-in light mechanism 12 to slide. The first screw rod 141 cooperates with the camera slide 113 to drive the camera slide 113 to move in a first direction; the second screw rod 142 cooperates with the fill-in light slide 121 to drive the fill-in light slide 121 to move in the first direction.

[0075] The first flip bracket 152 is mounted on the top of the first push rod base 158. The first push rod base 158 can be a linear motor that can drive the first flip bracket 152 to move in a first direction. This makes it convenient to adjust the height of the fruit flip mechanism 15 according to the height of the fruit on the flexible fruit tray 5. The housing of the first worm gear motor 157 is fixedly connected to the first flip bracket 152. The first drive gear 155 and the first driven gear 156 are both located inside the first flip bracket 152. The first drive gear 155 is coaxially connected to the output shaft of the first worm gear motor 157. The first worm gear motor 157 is used to drive the first drive gear 155 to rotate. The first drive gear 155 and the first driven gear 156 are meshed and connected via a first transmission rack 154. The first drive gear 155 is used to drive the first transmission rack 154 to move along the second direction of the device. The second direction of the device is perpendicular to the conveying direction of the conveyor belt. The first rotating motor 153 is fixedly connected to the first flip bracket 152 through the motor frame. The output shaft of the first rotating motor 153 is coaxially connected to the first contact disk 151 or is driven by a belt. The output shaft of the first rotating motor 153 is parallel to the width direction of the device. The two oppositely arranged first contact disks 151 are used to clamp the fruit to be inspected, and the first rotating motor 153 is used to achieve flipping.

[0076] In a specific implementation, the lens center of the industrial camera 112, the center of the inner fill light ring 123, and the center of the optical fiber probe of the spectrometer are coaxially arranged in the vertical direction; the axis of the first sensing mechanism 16, the first contact plate 151 and the inner fill light ring 123 are arranged in the same plane along the plane where the DR2 direction - DR1 direction is located.

[0077] <Hardness Information Collection Area 2>

[0078] like Figure 14-16 As shown, the hardness information collection area 2 is equipped with a hardness collection mechanism 21, which includes a rotatable hardness collection rotating platform 211. The hardness collection rotating platform 211 is fixedly mounted with a fruit hardness detection mechanism 212 and a fruit surface support mechanism 213 positioned opposite each other. The present invention utilizes a conveyor mechanism 4 to transport the fruit to the hardness collection mechanism 21 for collection of the fruit's hardness information. The rotatable hardness collection rotating platform 211 can collect hardness information from multiple angles, providing more accurate hardness information and avoiding errors caused by manual operation.

[0079] The hardness collection mechanism 21 also includes a second frame 214, a third screw 215, and a second sensing mechanism 216. The third screw 215 is rotatably connected to the upper portion of the second frame 214. The third screw 215 is externally connected to a screw motor, which can drive the third screw 215 to rotate around its own axis, driving the hardness collection rotary table 211 to move in the third direction. The hardness information collection area 2 is provided with a second sensing mechanism 216, which is fixedly connected to the second frame 214. The second sensing mechanism 216 is configured to sense whether the flexible fruit tray 5 enters a specific position in the hardness information collection area 2. Specifically, the second sensing mechanism 216 is two oppositely arranged photoelectric sensor switches, which are fixedly mounted on the main frame 41 of the transmission mechanism. The two photoelectric sensor switches are used to monitor the stop position of the flexible fruit tray 5 in the hardness information collection area 2.

[0080] The hardness collection rotary platform 211 includes a hardness slide 221 and a rotating platform 222 embedded within the hardness slide 221. A second pulley motor 223 is mounted on the hardness slide 221. The output shaft of the second pulley motor 223 is connected to the inner ring drive belt of the rotating platform 222 via a second small pulley 224, driving the rotating platform 222. The rotating platform 222 is fixedly mounted with at least one pair of opposed fruit hardness detection mechanisms 212 and fruit surface support mechanisms 213. The fruit hardness detection mechanisms 212 and fruit surface support mechanisms 213 are each connected to the second frame 214 via the hardness collection rotary platform 211.

[0081] The hardness slide 221 is connected to the third screw rod 215 so as to be movable up and down. The rotating table 222 is coaxially installed on the hardness slide 221 through a ring guide rail and can rotate around its own axis. The axis of the rotating table 222 is parallel to the vertical direction. The fruit hardness detection mechanism 212 and the fruit surface support mechanism 213 are respectively installed on both sides of the rotating table 222. The second small pulley 224 is connected to the outer periphery of the rotating table 222 through a belt. The second small pulley 224 is externally connected to the second pulley motor 223. The second pulley motor 223 is used to drive the second small pulley 224 to rotate, and then drive the rotating table 222 to rotate around its own axis through belt friction.

[0082] The fruit hardness detection mechanism 212 includes a hardness detection probe 231 and a second pushing mechanism 230 secured to the probe 231. The second pushing mechanism 230 is configured to push the probe 231 in a second direction to contact the fruit surface. The present invention utilizes the hardness collection mechanism 21 on a rotatable rotating platform 222, enabling the collection of hardness information from different parts of persimmon fruits, further enhancing the comprehensiveness and accuracy of the measurement.

[0083] The second driving mechanism 230 includes a second tilt bracket 232, a second transmission rack 233 disposed within the second tilt bracket 232, and a second worm gear motor 234. The second transmission rack 233 is engaged with a second drive gear 235 and a second driven gear 236, respectively. The second drive gear 235 and the second driven gear 236 are mounted within the second tilt bracket 232. The second drive gear 235 is connected to the output shaft of the second worm gear motor 234, driving the second transmission rack 233 to move horizontally. The hardness detection probe 231 is fixedly mounted at the end of the second transmission rack 233.

[0084] In the second propulsion mechanism 230, a second tilting bracket 232 is fixedly connected to the bottom end of the rotating platform 222. The housing of the second worm gear motor 234 is fixedly connected to the second tilting bracket 232. A second drive gear 235 and a second driven gear 236 are both located within the second tilting bracket 232. The second drive gear 235 is coaxially connected to the output shaft of the second worm gear motor 234. The second drive gear 235 and the second driven gear 236 are meshed and driven by a second transmission rack 233, which is arranged horizontally. The second drive gear 235 and the second driven gear 236 have approximately the same outer diameter. The second worm gear motor 234 drives the second drive gear 235 to rotate, thereby driving the second transmission rack 233 in the second direction. A hardness detection probe 231 is disposed at one end of the second transmission rack 233, which allows the hardness detection probe 231 to contact the surface of the fruit, thereby collecting hardness information.

[0085] The fruit surface supporting mechanism 213 includes a fruit surface supporting pad 241 and a third pushing mechanism 240 . The third pushing mechanism 240 is configured to push the fruit surface supporting pad 241 to move along the second direction to contact the fruit surface.

[0086] The third pushing mechanism 240 includes a third flip bracket 242, a third transmission rack 243 arranged in the third flip bracket 242, and a third worm gear motor 244. The third transmission rack 243 is respectively engaged with a third driving gear 245 and a third driven gear 246. The third driving gear 245 and the third driven gear 246 are installed in the third flip bracket 242, and the third driving gear 245 is connected to the output shaft of the third worm gear motor 244 to drive the third transmission rack 243 to move in the horizontal direction.

[0087] The fruit surface support pad 241 is fixedly mounted on the end of the third transmission rack 243 and has a concave surface that matches the fruit surface.

[0088] The third flip bracket 242 in the third pushing mechanism 240 is fixedly connected to the bottom end of the rotating table 222, the outer shell of the third turbine-worm motor 244 is fixedly connected to the third flip bracket 242, the third driving gear 245 and the third driven gear 246 are both located inside the third flip bracket 242, and the third driving gear 245 is coaxially connected to the output shaft of the third turbine-worm motor 244, the third driving gear 245 and the third driven gear 246 are meshed and connected through the third transmission rack 243, the third transmission rack 243 is horizontally arranged, and a fruit surface support pad 241 is provided at one end of the third transmission rack 243, and the hardness detection probe 231 and the fruit surface support pad 241 are relatively coaxially arranged; the second turbine-worm motor 234 drives the second transmission rack 233 to move horizontally, and the third turbine-worm motor 244 drives the third transmission rack 243 to move horizontally, so that after the hardness detection probe 231 and the fruit surface support pad 241 clamp the fruit, the fruit is punctured to collect hardness information.

[0089] <Sugar Content Information Collection Area 3>

[0090] like Figures 17-19 As shown, the sugar content information collection area 3 is provided with a sugar content collection mechanism 31, which includes a fruit transfer mechanism 32, an automatic juice extraction mechanism 33, and a sugar content measurement mechanism 34. The fruit transfer mechanism 32 is configured to flip the fruit from the flexible fruit tray 5 and transfer it to the juicing pot of the automatic juice extraction mechanism 33; the automatic juice extraction mechanism 33 is configured to crush and filter the fruit to obtain juice; and the sugar content measurement mechanism 34 is configured to measure the sugar content of the fruit juice. The present invention combines the fruit transfer mechanism 32, the automatic juice extraction mechanism 33, and the sugar content measurement mechanism 34 to achieve a fully automated process from fruit juicing to sugar content measurement. This eliminates the need for separate cleaning of the device after each sugar content collection, thereby improving the usability and automation level of the equipment.

[0091] The sugar content collecting mechanism 31 further includes a support frame, on which the sugar content measuring mechanism 34 is mounted. The automatic juice extracting mechanism 33 and the fruit transferring mechanism 32 are both mounted on a main frame 41 of the conveying mechanism 4 .

[0092] The fruit transfer mechanism 32 is used to transfer the fruit on the flexible fruit tray 5 to the juice pot of the automatic juice extraction mechanism 33. The fruit transfer mechanism 32 can be structurally identical or equivalent to the fruit flipping mechanism 15 within the image spectrum acquisition area 1, and will not be described in detail here. Typically, as shown in the figure, the fruit transfer mechanism 32 includes a first contact plate 151 and a first pushing mechanism 150 fixed to the transmission mechanism. The first contact plate 151 is mounted at one end of the first pushing mechanism 150 and is configured to contact and position the persimmon fruit. The first pushing mechanism 150 is configured to push the persimmon fruit on the flexible fruit tray 5 into the juice pot of the automatic juice extraction mechanism 33.

[0093] The sugar content information collection area 3 is further provided with a third sensing mechanism 35, which is configured to sense whether the flexible fruit tray 5 has entered a specific position in the sugar content information collection area 3. Specifically, the third sensing mechanism 35 comprises two opposing photoelectric sensors fixedly mounted on the main frame 41 of the transmission mechanism. The two photoelectric sensors are used to monitor the stopping position of the flexible fruit tray 5 in the sugar content information collection area 3.

[0094] Typically, there are two fruit transfer mechanisms 32 , which are arranged opposite to each other in the second direction of the transmission mechanism. The two fruit transfer mechanisms 32 are respectively fixedly connected to both sides of the main frame 41 in the transmission mechanism 4 .

[0095] The automatic juicing mechanism 33 includes a juicing pot body 331 for accommodating fruits, a fruit guide groove 332 provided on one side of the juicing pot body 331, and a fruit crushing mechanism provided at the upper end of the juicing pot body 331; a residue processing knife 333 is provided along the inner wall of the juicing pot body 331, and the blade of the residue processing knife 333 has a curvature equivalent to that of the inner wall of the juicing pot body 331, and a processing knife driving motor 334 for driving the residue processing knife 333 is provided on the outside of the juicing pot body 331, and the processing knife driving motor 334 drives the residue processing knife 333 to remove the residue in the juicing pot body 331; a plurality of filter holes are provided at the bottom of the juicing pot body 331.

[0096] The fruit crushing mechanism includes a juicer knife 335, a juicer motor 336 that drives the juicer knife 335 to perform a crushing operation, and a juicer knife position adjustment mechanism that transports the juicer knife 335 to the juicer pot body 331. The juicer knife position adjustment mechanism includes two second push rod seats 337, and the movable ends of the two second push rod seats 337 are connected by a connecting rod structure 338, and the connecting rod structure 338 is fixed to the juicer motor 336; the two second push rod seats 337 push the connecting rod structure 338 to move, and move the telescopic juicer knife 335 into the juicer pot body 331 to perform a juicing operation.

[0097] like Figure 5 and Figure 17 As shown, two second push rod seats 337 are fixedly connected to either side of the main frame 41 of the conveying mechanism 4. The ends of a connecting rod structure 338 are respectively connected to the bottom ends of the two second push rod seats 337 and can move up and down as the two second push rod seats 337 are driven. The housing of the juice extraction motor 336 is fixedly connected to the middle of the connecting rod structure 338. The output shaft of the juice extraction motor 336 is coaxially connected to the juice extraction knife 335, thereby driving the juice extraction knife 335 to rotate and crush the fruit. The juice extraction knife 335 is used to crush the fruit and extract the juice.

[0098] like Figure 5 and Figure 18As shown, the juicing pot body 331 is fixedly mounted on the main frame 41 of the conveying mechanism 4, the shell of the processing knife driving motor 334 is fixedly connected to the outer wall of the juicing pot body 331, and the output shaft of the processing knife driving motor 334 is connected to the residue processing knife 333. The residue processing knife 333 is used to process the residue after juicing, and the residue is scraped out from the openings on both sides of the juicing pot body, and the cleaning operation is carried out using the flushing water pipes 339 on both sides; the bottom of the juicing pot body is provided with a filter hole; when the automatic juicing mechanism 33 is used to squeeze juice from the fruit, the sugar content measuring mechanism 34 is located directly below the juicing pot body, and the juicing knife 335 is located inside the juicing pot body.

[0099] The sugar content measuring mechanism 34 is a refractometer measuring mechanism for measuring the sugar content of the fruit. Specifically, the sugar content measuring mechanism 34 includes a plurality of refractometer measuring devices 341, a first refractometer driving mechanism 342 that sequentially drives the plurality of refractometer measuring devices 341 to the bottom of the juice extraction pot 331, and a second refractometer driving mechanism 343 that drives the corresponding refractometer measuring devices 341 to flip. The refractometer measuring device 341 includes a funnel 318 fixed to the second refractometer driving mechanism, and a circular filter 319 is provided inside the funnel 318. , its bottom end is fixedly connected to the refractometer 320; the second refraction driving mechanism 343 includes a first bevel gear 344, a second bevel gear 345 corresponding to each refractometer measuring device 341, and a bevel gear driving motor 346 for driving the first bevel gear 344 to rotate, wherein the central axis of the second bevel gear 345 is fixedly connected to the funnel 318, the first bevel gear 344 is engaged with each second bevel gear 345, and when the bevel gear driving motor 346 drives the first bevel gear 344 to rotate, the funnel 318 performs a flipping operation.

[0100] The first refractive index drive mechanism 342 includes an electric push rod 351, a rotary drive motor 352, and a bevel gear structure. The bevel gear structure is sleeved on the outer periphery of the push rod shaft of the electric push rod 351. The rotary drive motor 352 drives the bevel gear structure to rotate so that each refractometer 341 is transported to the bottom of the juice pot body 331 in the order of rotation. The electric push rod seat of the electric push rod 351 is vertically fixed on the support frame, and its push rod shaft is coaxially connected to the upper part of the electric push rod seat so as to be movable up and down. The bevel gear structure is vertically sleeved on the outer periphery of the push rod shaft so as to be rotatable along its own axis. The bevel gear structure is fixedly connected to the outer shell of the rotation drive motor 352, and the output shaft of the rotation drive motor 352 is connected to the electric push rod shaft through a belt transmission. The outer shell of the bevel gear drive motor 346 is connected to the bevel gear structure, and the output shaft of the bevel gear drive motor 346 is coaxially connected to the first bevel gear 344. The three second bevel gears 345 are all meshed with the first bevel gear 344 for transmission. The inner periphery of each first bevel gear 344 is fixedly sleeved with a straight rod, and the end of the straight rod is fixedly connected to the funnel 318. The inner wall surface and the bottom end of the funnel 318 are coaxially fixedly connected to the circular filter 319 and the refractometer 320 respectively.

[0101] Because the output shaft of the rotary drive motor 352 is connected to the electric push rod shaft via a belt drive, and the electric push rod shaft can only move up and down and cannot rotate about its own axis, when the output shaft of the rotary drive motor 352 rotates, the reaction force exerted by the electric push rod shaft on the rotary drive motor 352 causes the rotary drive motor 352 to orbit around the electric push rod shaft, thereby driving the bevel gear structure to rotate relative to the electric push rod shaft. The axial direction of the first bevel gear 344 is parallel to the vertical direction, and the axial direction of the second bevel gear 345 is parallel to the radial direction of the electric push rod shaft.

[0102] <Computer Equipment Section>

[0103] like Figure 20 As shown, it is a principle block diagram of the computer device of the present invention. The computer device can be an electronic device, such as a personal computer (PC), a server device, a mobile device, an embedded device, etc. This electronic device performs model modeling based on the information provided by other components in the system. The electronic device may include a processor, RAM, a neural network chip, a memory, a sensor, and a communication module. It may also include an input / output module, a security module, a power control device, etc. The processor can control the overall operation of the electronic device. The processor may include one processor core (single-core) or multiple processor cores (multi-core). The processor can process or execute programs and / or data stored in the memory. In some example embodiments, the processor can control the functions of the neural network chip by executing programs stored in the memory. The processor can be implemented as a CPU, a GPU, an APU, etc.

[0104] The memory may include random access memory (RAM) (such as dynamic random access memory (DRAM) and static random access memory (SRAM)), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), CD-ROM, Blu-ray or other optical disk storage devices, hard disk drives (HDDs), solid-state drives (SSDs), or flash memory. The device can analyze input data in real time based on the system to extract valid information and implement a fully automated fruit quality evaluation model based on the extracted information. The memory is a storage location for storing data and can store an operating system (OS), various programs, and various data.

[0105] The sensor can collect peripheral information of the electronic device on which the electronic device is mounted. The sensor can sense or receive signals (e.g., video signals, audio signals, magnetic signals, bio-signals, touch signals, etc.) from outside the electronic device and convert the sensed or received signals into data. For example, the sensor can include at least one of various types of sensing devices (such as a microphone, an imaging device, an image sensor, a light detection and ranging (LIDAR) sensor, an ultrasonic sensor, an infrared sensor, a biosensor, and a touch sensor). The sensor can be the first sensing mechanism 16, the second sensing mechanism 216, and the third sensing mechanism 35 provided on the transmission mechanism.

[0106] The communication module may be provided with various wired or wireless interfaces capable of communicating with external devices. For example, the communication module may include a communication interface capable of accessing a wired local area network (LAN), a wireless local area network (WLAN) (such as Wireless Fidelity (Wi-Fi)), a wireless personal area network (WPAN) (such as Bluetooth), a wireless universal serial bus (Wireless USB), Zigbee, near field communication (NFC), radio frequency identification (RFID), power line communication (PLC), or a mobile cellular network (such as third generation (3G), fourth generation (4G), and long term evolution (LTE)).

[0107] The above system can be used to build a fully automated fruit quality evaluation model. The specific method includes the following steps:

[0108] (S1) placing the fruit to be inspected on the flexible fruit tray 5, and using the conveying mechanism 4 to convey the fruit to the image spectrum acquisition area 1, and performing all-round lossless acquisition of the image information and spectrum information of the fruit;

[0109] (S2) using the conveying mechanism 4 to sequentially convey the fruit to the hardness information collection area 2 and the sugar content information collection area 3, and sequentially collecting the hardness information and sugar content information of the fruit;

[0110] (S3) Constructing a prediction model for the internal and external quality of the fruit based on the image information, spectral information, hardness information and sugar content information of the fruit.

[0111] Wherein step (S1) specifically comprises:

[0112] (S11) placing the fruit to be tested on the flexible fruit tray 5, and the conveying mechanism 4 drives the flexible fruit tray 5 to move to the image spectrum collection area 1;

[0113] (S12) dimming and filling light in the dark box 14, splitting the light beam emitted by the filling light source by the spectrometer 118, and collecting and obtaining the transmission spectrum information of the fruit after filling light only from the spectral filling light station;

[0114] (S13) light is adjusted and filled in the dark box 14, and the light beam emitted by the fill light source is split by the spectrometer 118, and the industrial camera 112 collects and obtains the surface image information of the fruit after the fill light is filled in only from the camera fill light station;

[0115] (S14) The fruit is flipped over to expose different fruit surfaces, and steps (S12) to (S13) are repeated to collect transmission spectrum information and surface image information of different fruit surfaces of the fruit to be tested.

[0116] When the present invention is implemented, the specific steps are as follows: first, the persimmon fruit to be tested is placed on the flexible fruit tray 5, and the fruit is transported to the dark box 14 by the conveying mechanism 4. The two conveyor belts in the conveying mechanism 4 drive the flexible fruit tray 5 to move to the bottom of the fill light mechanism 12 in the dark box 14; then the halogen light source 129 in the dark box 14 is started and the inner fill light ring 123 is rotated, so that the light beam emitted by the lamp head 116 is split by the spectrometer 118 and is only emitted from the spectral fill light U-shaped hole 127 to the current position of the surface of the persimmon fruit. At this time, the probe of the spectrometer collects the persimmon. Then, the halogen light source 129 in the darkroom 14 is started and the inner fill light ring 123 is rotated so that the light beam emitted by the lamp head 116 is split by the spectrometer 118 and then emitted only from the camera fill light U-shaped hole 128. After diffuse reflection by the inner surface of the light shielding plate 143, it is projected to the current position of the surface of the persimmon fruit. At this time, the industrial camera 112 collects the surface image information of one side of the persimmon fruit. The fruit turning mechanism 15 is used to clamp and turn the fruit, and then the above steps are repeated to collect the transmission spectrum information and surface image information of the other side of the persimmon fruit.

[0117] Specifically, persimmon fruits are placed on the flexible fruit tray 5, and the two conveyor belts synchronously drive the flexible fruit tray 5 to move along the extension direction of the conveyor belts to the position to be detected in the dark box 14, and the distance information from the top of the current persimmon fruit is obtained through the distance sensor 111, and the distance between the camera slide 113 and the filling slide 121 and the top of the current persimmon fruit along the first direction is adjusted accordingly until a pre-set distance suitable for detection is reached, the halogen light source 129 is started and the inner filling light ring 123 is rotated, and the probe of the spectrometer is used to receive and store the transmission spectrum of the persimmon fruit, and the inner filling light ring 123 is repeatedly rotated until the light emitted by the lamp head 116 traverses all the spectral filling light U-shaped holes 127, and the inner filling light ring 123 is rotated to reset the halogen light source 129; then, the halogen light source 129 is started and the inner filling light ring 123 is rotated until the light emitted by the lamp head 116 passes through the spectrometer 118 and only comes from the camera filling light ring 123. The light is emitted from the U-shaped hole 128 to the position on the surface of the light shielding plate 143. At this time, the industrial camera 112 captures the image information of the fruit surface, and the rotating inner fill light ring 123 resets the halogen light source 129; then the first transmission rack 154 of the fruit turning mechanism 15 drives the first contact disk 151 to be adjusted along the width direction to just contact the fruit surface, and the first push rod pedestal 158 of the relatively arranged fruit turning mechanism 15 is started at the same time, and the fruit is driven to move along the first direction to separate from the flexible fruit tray 5 through the first turning bracket 152. The first rotating motor 153 is started at the same time and drives the pair of first contact disks 151 and the fruit to rotate 180° through the belt drive, and then drives the fruit to move along the first direction to contact the flexible fruit tray 5 through the first turning bracket 152, and the fruit turning mechanism 15 is reset; repeat the above steps, and the transmission spectrum and image information collection of the current persimmon fruit are completed.

[0118] Step (S2) is specifically:

[0119] (S21) using the two conveyor belts in the conveying mechanism 4 to drive the flexible fruit holder 5 to move into the hardness collecting mechanism 21, and using the fruit hardness detecting mechanism 212 and the fruit surface supporting mechanism 213 in the hardness collecting mechanism 21 to clamp the fruit and puncture the fruit to collect the hardness information of the fruit;

[0120] (S22) Then, the two conveyor belts in the conveying mechanism 4 are used to drive the flexible fruit tray 5 to the position where the sugar content collecting mechanism 31 is located, and the funnel 318 in the refractometer measuring mechanism is moved to the position directly below the juice extractor;

[0121] (S23) Then, the fruit is clamped by the fruit transfer mechanism 32 in the sugar content collection mechanism 31 and thrown into the juicer. The fruit is juiced by the automatic juicer 33. The squeezed juice enters the refractometer measuring mechanism from the juicer, and the sugar content information of the juice is collected by the refractometer measuring mechanism.

[0122] During the specific implementation, the two conveyor belts synchronously drive the flexible fruit tray 5 to move along the extension direction of the conveyor belts to the position to be detected in the hardness collection mechanism 21, adjust the distance of the hardness slide 221 along the first direction to a pre-set distance suitable for detection, and the third transmission rack 243 of the fruit surface support mechanism 213 drives the fruit surface support pad 241 to adjust along the width direction to just contact the fruit surface. The second transmission rack 233 of the fruit hardness detection mechanism 212 drives the hardness detection probe 231 to adjust along the width direction to just contact the fruit surface and further press into the fruit to obtain the fruit skin and flesh hardness data; after the second and third transmission racks 243 are reset, the rotating table 222 is rotated to different positions multiple times to obtain the hardness information of the fruit at different positions. At this point, the hardness information collection of the current persimmon fruit is completed.

[0123] The two conveyor belts synchronously drive the flexible fruit tray 5 along the conveyor belt extension direction to the position to be detected in the sugar content collection mechanism 31. The first transmission rack 154 in the fruit transfer mechanism 32 drives the first contact plate 151 to adjust along the width direction until it just contacts the surface of the fruit. The oppositely arranged fruit transfer mechanisms 32 are activated simultaneously. The first flip bracket 152 drives the fruit in the vertical direction until it is free of the flexible fruit tray 5. The two conveyor belts synchronously drive the flexible fruit tray 5 in the length direction until the fruit is free of the flexible fruit tray 5. The first flip bracket 152 of the fruit transfer mechanism 32 drives the fruit in the vertical direction to the position closest to the juicing pot, and the fruit transfer mechanism 32 is reset. After the fruit has completely entered the juicer, the juicer 335 driven by the juicer motor 336 is synchronously driven by the second push rod seat 337 to break the fruit. The juice produced by the fruit drips from the filter holes at the bottom of the juicer through the circular filter 319 to the surface of the refractometer 320. After the device is completely stable, the refractometer 320 records the current sugar content information of the juice. At this point, the collection of the sugar content information of the current persimmon fruit is completed. Finally, a cleaning water pipe 339 is provided in the juice pot to spray water for cleaning, and cooperates with the residue processing knife 333 to clean the large pieces of fruit residue remaining in the juice pot. The water pipe 321 in the funnel 318 sprays water to clean the residual juice on the surface of the refractometer 320 and the circular filter 319. At the same time, the electric push rod seat drives the refractometer measuring mechanism away from the juice pot in the first direction, and starts the rotation drive motor 352 and the bevel gear drive motor 346, so that the funnel 318 rotates around the bevel gear structure and revolves to the next measurement position, preparing for the next fruit information measurement; after collecting a sufficient amount of fruit spectrum, image, hardness and sugar content information, a spectral analysis method combined with an image processing method is used to construct an internal and external quality prediction model of the fruit.

[0124] The position of the fruit to be detected is controlled by the first sensing mechanism 16, the second sensing mechanism 216 and the third sensing mechanism 35. When at each position to be detected, the fruit and the sensors used by these sensing mechanisms (such as photoelectric sensing switches) can be coaxial.

[0125] The specific steps of applying the spectral analysis method of the economic fruit automated quality evaluation model construction system of this embodiment are as follows:

[0126] (S41) The key equipment of the economic fruit and vegetable automated quality evaluation model construction system, such as the industrial camera, spectrometer, hardness acquisition mechanism 21, sugar content acquisition mechanism 31, first sensing mechanism 16, second sensing mechanism 216, and third sensing mechanism 35, are connected to the computer equipment, and each acquisition mechanism, sensing mechanism, industrial camera, spectrometer and other detection equipment are operated for a period of time to reach a stable state, so as to ensure the stability of the operating environment of the measurement system.

[0127] (S42) Using the economic fruit and forestry automated quality evaluation model construction system to collect reference spectra and dark field spectra in advance, turning off the light source when collecting the dark field spectrum, and then using the economic fruit and forestry automated quality evaluation model construction system to collect internal and external quality information of persimmon fruits from multiple angles.

[0128] (S43) Smoothing, detrending, and performing standard normal variable transformation (SNV) on the collected internal and external quality information data to reduce noise and interference signals and improve data quality.

[0129] (S44) Dimensionality reduction methods such as partial least squares regression (PLS) were used to compress high-dimensional spectral data into latent variables, screen out latent variables closely related to the physical and chemical properties of persimmon, and eliminate variables with large noise.

[0130] (S45) The data set is divided by a sample set partitioning method based on joint XY distances (SPXY), the bands are optimized by a moving window method, model parameters are set, k-fold cross-validation is performed, the optimal number of latent variables is selected for modeling, and model evaluation indicators are calculated to optimize the model, thereby realizing the construction of an automated prediction model for fruit internal and external quality information.

[0131] This embodiment uses persimmon fruit as an application for actual measurement. The fruit spectral information obtained in the embodiment is as follows: Figure 21 As shown, the hardness information obtained in the embodiment is as follows Figure 22 As shown, the sugar content information obtained in the embodiment is as follows Figure 23 As shown, the image information obtained by the embodiment is as follows Figure 24 As shown in the figure, the box selected in the figure is the location of the surface defect of the fruit. The hardness modeling result obtained in the embodiment is as follows Figure 25As shown in the example, the sugar content modeling results obtained are as follows Figure 26 As shown in Figure 2, the prediction model for the soluble sugar content in persimmon samples is R 2 The prediction accuracy reaches 87.2%, and the root mean square error (RMSEP) is 0.432°Bx. The prediction model predicts the hardness value of persimmon samples R 2 The prediction accuracy reached 45.3%, with a root mean square error (RMSEP) of 0.291N. This prediction model can identify surface defects and determine the external quality of persimmons. The present invention utilizes multiple, automated tests of persimmon samples across multiple dimensions to obtain spectral, hardness, sugar content, and image information. This allows for the accurate and automated construction of a prediction model for the internal and external quality of persimmons, providing a technical foundation for intelligent grading and quality assessment of persimmon-related fruits.

[0132] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

Claims

1. A system for constructing an automated quality evaluation model for economic fruit trees, characterized in that: Its settings are: An image spectrum acquisition area is configured to non-destructively acquire image information and spectral information of fruits; a hardness information collection area configured to collect hardness information of fruits; The sugar content information collection area is configured to collect sugar content information of fruits; The device also includes a conveying mechanism and a computer device. The conveying mechanism is provided with a flexible fruit tray for carrying fruit. The image spectrum collection area, the hardness information collection area, and the sugar content information collection area are sequentially arranged along the conveying direction of the conveying mechanism. The conveying mechanism conveys the fruit to the image spectrum collection area, the hardness information collection area, and the sugar content information collection area according to preset instructions to collect image information and spectrum information, hardness information, and sugar content information in sequence. The computer device constructs an automatic fruit quality evaluation model based on the collected image information, spectral information, hardness information and sugar content information.

2. The economic fruit tree automated quality evaluation model construction system according to claim 1 is characterized in that: The conveying mechanism includes a main frame, a conveyor belt assembly arranged in the main frame, and a conveying motor fixedly installed on the main frame. The flexible fruit tray is moved by the conveyor belt assembly to the image spectrum collection area, the hardness information collection area, and the sugar content information collection area according to preset instructions; Preferably, the conveyor belt assembly includes two conveyor belts arranged in parallel and a plurality of conveyor guide wheels arranged in sequence and at intervals along the conveying direction of the conveying mechanism; the conveyor guide wheels are mounted on the main frame and include a first conveyor wheel and a second conveyor wheel coaxially connected and arranged, one conveyor belt is sequentially sleeved on each first conveyor wheel, and the other conveyor belt is sequentially sleeved on each second conveyor wheel, and the conveyor motor drives the two conveyor belts to move synchronously according to preset instructions; Preferably, the output shaft of the conveying motor is connected to one of the several conveying guide wheels, and the two ends of the flexible fruit tray are respectively placed on the two conveyor belts and driven by the two conveyor belts.

3. The economic fruit tree automated quality evaluation model construction system according to claim 2 is characterized in that: The image spectrum acquisition area is provided with a dark box, and an image acquisition mechanism, a spectrum acquisition mechanism, and a fill light mechanism are provided in the dark box. The fill light mechanism is provided on a side of the flexible fruit tray close to the image acquisition mechanism, and the spectrum acquisition mechanism is provided on a side of the flexible fruit tray away from the image acquisition mechanism. Preferably, the image acquisition mechanism and the spectrum acquisition mechanism are arranged opposite to each other in a first direction, the fill light mechanism is located between the image acquisition mechanism and the spectrum acquisition mechanism, and the image capture path of the image acquisition mechanism, the light source supplement path of the fill light mechanism, and the spectrum acquisition path of the spectrum acquisition mechanism overlap; Preferably, when performing image spectrum acquisition, the image acquisition mechanism, the fill light mechanism, the spectrum acquisition mechanism, and the flexible fruit holder are coaxially arranged in a first direction; Preferably, the image spectrum acquisition area is further provided with a fruit turning mechanism, and the fruit turning mechanism is configured to turn over the fruit placed on the flexible fruit tray so as to expose different surfaces of the fruit; Preferably, the number of the fruit turning mechanisms in the image spectrum acquisition area is two, and the two fruit turning mechanisms are arranged opposite to each other in the second direction; Preferably, the fruit turning mechanism includes a first contact plate and a first pushing mechanism fixed to the transmission mechanism, wherein the first contact plate is mounted at one end of the first pushing mechanism and is configured to contact the fruit and position the fruit; the first pushing mechanism is configured to push the fruit on the flexible fruit tray to change the exposed surface of the fruit; Preferably, the image spectrum collection area is further provided with a first sensing mechanism, and the first sensing mechanism is configured to sense whether the flexible fruit tray enters a certain position in the image spectrum collection area; Preferably, the first pushing mechanism includes a first flipping bracket and a first rotating motor, wherein the first rotating motor is mounted on the first flipping bracket and is configured to drive the first contact plate to rotate so as to flip the fruit; Preferably, a first transmission rack is provided in the first flip bracket, wherein the lower end of the first transmission rack is respectively engaged with a first driving gear and a first driven gear of the same outer diameter, and one end of the first transmission rack is fixed to the first contact plate, the first driving gear and the first driven gear are installed in the first flip bracket, and the first driving gear is connected to the output shaft of the first worm gear motor to drive the first transmission rack to move in the second direction; Preferably, the lower end of the first flip bracket is mounted and fixed to the main frame of the conveying mechanism through a first push rod base.

4. The economic fruit tree automated quality evaluation model construction system according to claim 3 is characterized in that: The fill light mechanism includes a movable fill light table, an outer light shield fixed to the fill light table, and an inner fill light ring embedded in the inner side of the outer light shield; the outer light shield and the inner fill light ring are coaxially arranged, and a first pulley motor is provided on the outer light shield, and the output shaft of the first pulley motor is connected to the inner ring transmission belt sleeved on the inner fill light ring through a first small pulley, driving the inner fill light ring to rotate and switch between the camera fill light position and the spectrum fill light position; A plurality of halogen light sources are evenly distributed circumferentially on a side of the inner light-fill ring facing the transmission mechanism; the outer light-shielding cover includes an inner ring and an outer ring formed integrally; the halogen light source is arranged between the inner ring and the outer ring of the outer light-shielding cover; the inner ring of the outer light-shielding cover is provided with a plurality of U-shaped holes for spectral light-filling; the outer ring of the outer light-shielding cover is provided with a plurality of U-shaped holes for camera light-filling; the plurality of U-shaped holes for spectral light-filling and the plurality of U-shaped holes for camera light-filling are arranged alternately; The halogen light source includes a halogen light source housing, a lamp holder, a condenser and a spectrometer. The lamp holder and the condenser are coaxially connected to the inner side of the halogen light source housing. The spectrometer is connected to the inner bottom surface of the halogen light source housing. The spectrometer is used to split the light beam emitted by the halogen light source into two beams of light that are respectively irradiated to the outer circle and the inner circle of the outer light shield. According to the work station of the halogen light source, the light separated by the spectrometer is filled onto the fruit by a camera fill light U-shaped hole or a spectrum fill light U-shaped hole.

5. The economic fruit tree automated quality evaluation model construction system according to claim 2 is characterized in that: The image acquisition mechanism includes a distance sensor for sensing the distance of the fruit on the flexible fruit tray, an industrial camera for imaging the fruit to be detected, and a camera slide for adjusting the relative position of the industrial camera, wherein the industrial camera can move in a third direction; Preferably, the darkroom includes a darkroom body, a driving mechanism for driving an image acquisition mechanism and a fill light mechanism to slide inside the darkroom body; the camera slide and the fill light slide are connected to the driving mechanism to drive the image acquisition mechanism and the fill light mechanism to slide; Preferably, the driving mechanism includes a first screw rod and a second screw rod installed in the dark box body, the first screw rod cooperates with the camera slide to drive the camera slide to move along the third direction; the second screw rod cooperates with the compensating slide to drive the compensating slide to move along the third direction.

6. The economic fruit tree automated quality evaluation model construction system according to claim 1 is characterized in that: The hardness information collection area is provided with a hardness collection mechanism, which includes a rotatable hardness collection rotating platform, and the hardness collection rotating platform is fixedly installed with a fruit hardness detection mechanism and a fruit surface support mechanism that are relatively arranged; Preferably, the fruit hardness detection mechanism includes a hardness detection probe and a second pushing mechanism fixed to the hardness detection probe; the second pushing mechanism is configured to push the hardness detection probe to move along a second direction to contact the surface of the fruit; Preferably, the second pushing mechanism includes a second flip bracket, a second transmission rack arranged in the second flip bracket, and a second worm gear motor, the second transmission rack is respectively engaged with a second driving gear and a second driven gear, the second driving gear and the second driven gear are installed in the second flip bracket, and the second driving gear is connected to the output shaft of the second worm gear motor to drive the second transmission rack to move in the horizontal direction; The hardness detection probe is fixedly mounted on the end of the second transmission rack; Preferably, the fruit surface support mechanism includes a fruit surface support pad and a third pushing mechanism, wherein the third pushing mechanism is configured to push the fruit surface support pad to move along the second direction to contact the fruit surface; Preferably, the third pushing mechanism includes a third flip bracket, a third transmission rack arranged in the third flip bracket, and a third worm gear motor, the third transmission rack is respectively engaged with a third driving gear and a third driven gear, the third driving gear and the third driven gear are installed in the third flip bracket, and the third driving gear is connected to the output shaft of the third worm gear motor to drive the third transmission rack to move in the horizontal direction; The fruit surface support pad is fixedly mounted on the end of the third transmission rack and has a concave surface that matches the fruit surface; The hardness collection rotating platform includes a hardness slide and a rotating platform embedded inside the hardness slide. A second pulley motor is provided on the hardness slide. The output shaft of the second pulley motor is connected to the inner ring transmission belt sleeved on the rotating platform through a second small pulley to drive the rotating platform to rotate. At least one set of relatively arranged fruit hardness detection mechanisms and fruit surface support mechanisms is fixedly mounted on the rotating platform.

7. The economic fruit tree automated quality evaluation model construction system according to claim 1 is characterized in that: The sugar content information collection area is provided with a sugar content collection mechanism, which includes a fruit transfer mechanism, an automatic juice extraction mechanism, and a sugar content measurement mechanism. The fruit transfer mechanism is configured to flip the fruit from the flexible fruit tray to the juice extraction pot of the automatic juice extraction mechanism; the automatic juice extraction mechanism is configured to crush and filter the fruit to obtain juice; and the sugar content measurement mechanism is configured to measure the sugar content information of the fruit juice. Preferably, the automatic juicing mechanism comprises a juicing pot body for accommodating fruits, a fruit guiding groove provided on one side of the juicing pot body, and a fruit crushing mechanism provided on the upper end of the juicing pot body; a residue processing knife is provided along the inner wall of the juicing pot body, the blade of the residue processing knife has a curvature equivalent to that of the inner wall of the juicing pot body, and a knife driving motor for driving the residue processing knife is provided on the outer side of the juicing pot body, and the knife driving motor drives the residue processing knife to remove residue in the juicing pot body; a plurality of filter holes are provided at the bottom of the juicing pot body; Preferably, the fruit crushing mechanism includes a juicer, a juicer motor that drives the juicer to perform crushing operations, and a juicer position adjustment mechanism that transports the juicer into the juicer pot body. The juicer position adjustment mechanism includes two second push rod seats and a connecting rod structure connected to the movable ends of the two second push rod seats, and the connecting rod structure is fixed to the juicer motor; the two second push rod seats push the connecting rod structure to displace, and move the telescopic juicer into the juicer pot body to perform juicing operations. Preferably, the sugar content measuring mechanism includes a plurality of refractometers, a first refractive index driving mechanism that sequentially drives the plurality of refractometers to the bottom of the juice extractor body, and a second refractive index driving mechanism that drives the corresponding refractometers to flip over; the refractometer includes a funnel fixed to the second refractive index driving mechanism, a circular filter is provided on the inside of the funnel, and the bottom end of the filter is fixedly connected to the refractometer; the second refractive index driving mechanism includes a first bevel gear, a second bevel gear corresponding to each refractometer, and a bevel gear driving motor that drives the first bevel gear to rotate, wherein the central axis of the second bevel gear is fixedly connected to the funnel, the first bevel gear is meshed with each second bevel gear, and when the bevel gear driving motor drives the first bevel gear to rotate, the funnel performs a flipping operation; Preferably, the first refraction drive mechanism includes an electric push rod shaft, a rotary drive motor, and a bevel gear structure. The bevel gear structure is sleeved on the outer periphery of the electric push rod shaft. The rotary drive motor drives the bevel gear structure to rotate so that each refractometer is transported to the bottom of the juice pot in a rotational order.

8. A method for building a fruit quality evaluation model using the device according to any one of claims 1 to 7, characterized in that: The following steps are involved: (S1) placing the fruit to be inspected on a flexible fruit tray, and using a conveying mechanism to convey the fruit to an image spectrum acquisition area, and performing all-round lossless acquisition of image information and spectrum information of the fruit; (S2) using a conveying mechanism to sequentially convey the fruit to a hardness information collection area and a sugar content information collection area, and sequentially collecting the hardness information of the fruit and the sugar content information of the fruit; (S3) Constructing a prediction model for the internal and external quality of the fruit based on the image information, spectral information, hardness information and sugar content information of the fruit.

9. The method for modeling a fruit quality evaluation model according to claim 8, wherein: Wherein step (S1) specifically comprises: (S11) placing the fruit to be detected on the flexible fruit tray, and the conveying mechanism drives the flexible fruit tray to move to the image spectrum acquisition area; (S12) dimming and filling light in a dark box, splitting the light beam emitted by the filling light source by a spectrometer, and collecting and obtaining the transmission spectrum information of the fruit after filling light only from the spectral filling light station by a spectrometer; (S13) adjusting the light and filling the light in the dark box respectively, splitting the light beam emitted by the fill light source by a spectrometer, and collecting and obtaining the surface image information of the fruit after the fill light is filled only from the camera fill light station by an industrial camera; (S14) The fruit is flipped over to expose different fruit surfaces, and steps (S12) to (S13) are repeated to collect transmission spectrum information and surface image information of different fruit surfaces of the fruit to be tested.

10. Application of the economic forest fruit automated quality evaluation model construction system according to any one of claims 1 to 7 in building a fruit quality evaluation model.

Images