Asparagus preservation monitoring system based on machine vision

Abstract

The invention relates to the field of transportation control, in particular to a machine vision-based asparagus fresh-keeping monitoring system, which comprises a transfer box which is arranged to comprise a spliced box body with an open surface, box top covering cloth and a video collector which is used for shooting a top outline image of the corresponding box top covering cloth and a surface image of the corresponding side surface of the transfer box; the transfer center console is used for analyzing the top contour image and the surface image of the side face to judge the stability of the transfer box and sending out a corresponding alarm according to a judgment result, and the external analyzer is used for shooting a target asparagus path for transferring of the transfer box; by fusing multi-source information, the shake state of the transfer box in the carrying process is effectively detected, the target asparagus path is adjusted in time, and the detection and evaluation precision of the shake of the transfer box and the cargo scattering risk is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of transportation control, and particularly relates to a machine vision-based asparagus fresh-keeping monitoring system. BACKGROUND

[0002] Asparagus is a kind of high-nutrition vegetable, and its freshness and integrity need to be maintained during transportation. However, due to the particularity of the shape of asparagus, the traditional transportation method often cannot guarantee its stability, and is prone to cause the asparagus to scatter and be damaged. The traditional manual monitoring method is inefficient, and it is difficult to find and solve the problems of asparagus damage and scattering caused by bumping, collision and other reasons in the transportation process in real time, resulting in huge economic losses.

[0003] Chinese patent application publication No. CN118522108A discloses a cargo transportation in-transit monitoring device, relating to the technical field of safety monitoring and alarm. The monitoring device comprises a supporting mechanism and a monitoring assembly. The monitoring assembly comprises a response mechanism and a transmission mechanism. The supporting mechanism is provided with side baffles at both ends, and each side baffle is provided with a bottom plate at the bottom. A slanted plate is arranged between the two bottom plates. The outer side of one of the bottom plates is provided with an outer protruding plate. The top of the outer protruding plate is provided with a center column. One end of the transmission mechanism is connected to the surface of the center column. The monitoring device can quickly detect the deviation of the cargo in any direction with the help of the monitoring assembly. The alarm process can be triggered when the abnormal state of the cargo is detected with the cooperation of the transmission mechanism and the response mechanism. The positioning rod can also be used to timely investigate the tilting phenomenon of the cargo. The supporting mechanism at the bottom can be used to monitor and protect the leakage of liquid cargo during transportation, thereby improving the safety of cargo transportation. However, the existing technology has the following problems: the existing technology does not consider the unreasonable path planning of moving the stacked asparagus to the transfer area, which may cause frequent shaking during the moving process, increasing the risk of cargo scattering. The existing technology lacks an effective detection method for the shaking state of the transfer box during the handling process, and cannot timely adjust the target path. SUMMARY

[0004] Therefore, the present application provides a machine vision-based asparagus fresh-keeping monitoring system to overcome the problem that the existing technology cannot effectively detect the shaking state of the transfer box during the handling process, and cannot timely adjust the target asparagus path.

[0005] To achieve the above-mentioned purpose, the present application provides a machine vision-based asparagus fresh-keeping monitoring system, comprising:

[0006] The transfer box is provided as a spliced box body with an open face, and each part of the box body is provided with a corresponding sealing strip;

[0007] The box top cover is arranged on the transfer box and is sealed with each sealing strip of the transfer box.

[0008] a video collector arranged above the transfer box to capture a top profile image of a corresponding top cover and a surface image of a corresponding side surface of the transfer box;

[0009] a transfer control console connected to the video collector to analyze the top profile image and the surface image to determine the stability of the transfer box and to issue a corresponding warning according to the determination result;

[0010] an external analyzer arranged on the transfer device and connected to the transfer control console to capture a target asparagus path of the transfer box during transfer;

[0011] wherein the target asparagus path is a path from a target asparagus position to a predetermined transfer position of the transfer box.

[0012] Further, the transfer control console determines a characteristic path point according to the curvature of the target asparagus path, and determines a characteristic path segment based on the characteristic path point.

[0013] The video collector captures a video stream of the transfer box moving along the characteristic path segment, and captures a top profile image of the transfer box in the video stream at a preset time interval, and the transfer control console determines a profile size difference based on each top profile image to determine whether the transfer box is a wobble performance characteristic transfer box.

[0014] The transfer control console, in response to the transfer box being a wobble performance characteristic transfer box, captures surface images of two mutually opposite side surfaces of the wobble performance characteristic transfer box based on the video stream, and records the area of the dark area of each surface image for the first time, and

[0015] In response to the wobble performance characteristic transfer box moving along the target asparagus path to the predetermined transfer position, the area of the dark area of the two side surfaces is recorded again, and whether the wobble performance characteristic transfer box has a cargo scattering risk is determined according to the change of the area of the dark area of each surface image.

[0016] Further, the transfer control console determines a scattering warning position as a scattering warning position for the wobble performance characteristic transfer box with a cargo scattering risk according to the target asparagus path and the cargo scattering risk, and obtains target asparagus paths of a plurality of wobble performance characteristic transfer boxes with a cargo scattering risk, determines a scattering negative incentive path segment based on the occurrence frequency of the characteristic path segment in a plurality of target asparagus paths, and adjusts the curvature of the scattering negative incentive path segment according to the change of the area of the dark area.

[0017] Further, the external analyzer establishes a rectangular coordinate system on the horizontal ground when determining the characteristic path segment, and determines coordinates of the path points on the target asparagus path in the rectangular coordinate system;

[0018] The numerical differentiation method is used to calculate the curvature of each path point;

[0019] The path point corresponding to the maximum curvature is determined as the characteristic path point;

[0020] The path segment with a preset length centered on the characteristic path point is determined as the characteristic path segment.

[0021] Further, when determining the contour size difference, the transfer control center determines the maximum contour width value and the maximum contour length value based on each top contour image;

[0022] The standard deviation of the maximum contour width value is calculated according to the maximum contour width value in each top contour image, and the standard deviation of the maximum contour length value is calculated according to the maximum contour length value in each top contour image;

[0023] The contour size difference representation coefficient is calculated according to the standard deviation of the maximum contour width value and the standard deviation of the maximum contour length value;

[0024] The contour size difference representation coefficient is positively correlated with the standard deviation of the maximum contour width value, and the contour size difference representation coefficient is positively correlated with the standard deviation of the maximum contour length value.

[0025] Further, when determining whether the transfer box is a shaking performance characteristic transfer box, the transfer control center compares the calculated contour size difference representation coefficient with a preset contour size difference representation coefficient comparison value;

[0026] If the contour size difference representation coefficient is greater than the contour size difference representation coefficient comparison value, the transfer box is determined to be a shaking performance characteristic transfer box.

[0027] Further, the process of recording the dark area area of each surface image by the transfer control center includes:

[0028] The captured surface image is denoised;

[0029] The pixel point with a pixel brightness lower than a preset pixel brightness threshold is determined as a dark pixel point;

[0030] The region composed of dark pixel points is determined as a dark region by an edge detection algorithm, and the dark region area is calculated by a pixel counting method;

[0031] determine a surface image of the mutually opposite two sides of the shaking performance characteristic transfer box as a first surface image and a second surface image;

[0032] determine a difference between the dark area of the first surface image recorded for the first time and the dark area of the first surface image recorded for the second time as a first dark area difference value, and determine a difference between the dark area of the second surface image recorded for the first time and the dark area of the second surface image recorded for the second time as a second dark area difference value.

[0033] Further, the transfer control console determines whether the shaking performance characteristic transfer box has a risk of goods scattering when determining the shaking performance characteristic transfer box,

[0034] If the first dark area difference value and the second dark area difference value are opposite in sign, it is determined that the shaking performance characteristic transfer box has a risk of goods scattering.

[0035] Further, when the transfer control console determines the scattering negative incentive path segment, the characteristic path segment corresponding to the target asparagus path of the shaking performance characteristic transfer box with a risk of goods scattering is obtained, and the characteristic path segment with the highest occurrence frequency is determined as the scattering negative incentive path segment.

[0036] Further, the process of adjusting the curvature of the scattering negative incentive path segment by the transfer control console comprises:

[0037] The average value of the absolute value of the first dark area difference value and the absolute value of the second dark area difference value is calculated, and the curvature of the scattering negative incentive path segment is adjusted according to the average value, and the curvature is negatively correlated with the average value.

[0038] Compared with the prior art, the beneficial effects of the present application are that the present application determines the characteristic path segment by obtaining the target asparagus path of the transfer box during transfer, collects the video stream of the transfer box moving along the characteristic path segment, determines the contour size difference based on each top contour image to determine whether the transfer box is a shaking performance characteristic transfer box, in response to the shaking performance characteristic transfer box, records the dark area of each surface image for the first time, moves to a predetermined transfer position, and records the dark area for the second time to determine whether the shaking performance characteristic transfer box has a risk of goods scattering, the predetermined transfer position of the shaking performance characteristic transfer box with a risk of goods scattering is warned, the scattering negative incentive path segment is determined based on the occurrence frequency of the characteristic path segment in a plurality of target asparagus paths, and the curvature of the scattering negative incentive path segment is adjusted according to the change of the dark area, thereby realizing effective detection of the shaking state of the transfer box during the transfer process by fusing multi-source information of video stream and image data, and timely adjusting the target asparagus path, and improving the detection and evaluation precision of the shaking of the transfer box and the risk of goods scattering.

[0039] Further, the present application determines the characteristic path point through the curvature of the target asparagus path, and determines the characteristic path segment according to the characteristic path point, and those skilled in the art can understand that the greater the curvature of the transfer box moving path, the more likely the transfer box will have a tendency to sway due to the greater centrifugal force, and the higher the stacking height of the transfer box during transfer transportation, the higher the transfer box center of gravity, which makes the centrifugal force more likely to affect the transfer box, so the greater the curvature, the worse the structural stability of the transfer box, the present application calculates the curvature of each path point by numerical differentiation, determines the path point corresponding to the maximum curvature as the characteristic path point, and determines the characteristic path segment according to the characteristic path point, and further, the path segment in the target asparagus path that is easy to affect the transfer box is screened, and the detection and evaluation accuracy of the swaying state of the transfer box during the handling process is improved.

[0040] Further, the present application determines whether the transfer box is a swaying performance characteristic transfer box by determining the contour size difference, and those skilled in the art can understand that the centrifugal force received by the transfer box in the characteristic path segment with greater curvature will cause the overall structure of the transfer box to be distorted and deformed, losing the original structural stability, and the transfer box top contour collected intuitively reflects the unstable condition of the internal structure of the transfer box, the more obvious the length and width size change of the transfer box top contour, the greater the influence on the internal structural stability of the transfer box, and the worse the stability of the transfer box, the present application determines whether the transfer box is a swaying performance characteristic transfer box by determining the contour size difference, and further, the degree of swaying of the transfer box is intuitively quantified, and the detection and evaluation accuracy of the swaying and cargo scattering risk of the transfer box is improved.

[0041] Further, the present application determines whether the swaying performance characteristic transfer box has a cargo scattering risk through the change of the dark area of each surface image, and it can be understood that the accumulated transfer box has stacking voids due to the difference in asparagus shape, and the image collected on the side of the transfer box will have a local dark area due to the stacking voids, the present application calculates the difference value between the dark area of the side image of the transfer box during the target asparagus process and the dark area of the side image of the transfer box after completing the target asparagus, which represents the change of the stacking voids of the transfer box during the target asparagus process, and the size change of the stacking voids of the two opposite sides of the transfer box is opposite, which represents that the greater the change of the internal structure of the originally stable transfer box, the more likely it is to cause the transfer box to scatter, and the detection and evaluation accuracy of the swaying and cargo scattering risk of the transfer box is improved.

[0042] Further, the present application adjusts the curvature of the scattering negative incentive path segment according to the change of the dark area, and those skilled in the art can understand that the greater the difference value between the dark area in the side image of the transport box during the target asparagus process and the dark area in the side image of the transport box after the target asparagus is completed, the greater the shaking degree of the transport box, the greater the influence of the transport box on the feature path segment, and the curvature of the feature path segment needs to be reduced to improve the stability of the transport of the transport box, thereby realizing the timely adjustment of the target asparagus path. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 For the connection diagram of the asparagus preservation monitoring system based on machine vision of the embodiment of the present application, including:

[0044] Figure 2 For the step diagram of the asparagus preservation monitoring system based on machine vision of the embodiment of the present application, including:

[0045] Figure 3 For the logic flow chart of determining whether the transport box is a shaking performance characteristic transport box of the embodiment of the present application, including:

[0046] Figure 4 For the step diagram of recording the dark area of each surface image of the embodiment of the present application, including:

[0047] Figure 5 For the logic flow chart of determining whether the shaking performance characteristic transport box has the risk of goods scattering of the embodiment of the present application. DETAILED DESCRIPTION

[0048] In order to make the purpose and advantages of the present application more clear and obvious, the present application will be further described below in combination with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present application, and not to limit the present application.

[0049] The preferred embodiments of the present application will be described below with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are only used to explain the technical principles of the present application, and are not intended to limit the protection scope of the present application.

[0050] It should be noted that in the description of the present application, the terms "upper", "lower", "inner", "outer" and the like indicate the direction or positional relationship of the terms based on the direction or positional relationship shown in the drawings, which is only for the convenience of description, and does not indicate or imply that the device or element must have a particular orientation, be constructed and operated in a particular orientation, therefore it cannot be understood as a limitation of the present application.

[0051] Please refer to Figure 1 shown, which is the connection diagram of the asparagus preservation monitoring system based on machine vision of the embodiment of the present application, including:

[0052] The transport box is provided as a spliced box body including an open face, and each part of the box body is provided with a corresponding sealing strip;

[0053] The box top cover is arranged on the transport box and is sealed with each sealing strip of the transport box;

[0054] The video collector is arranged above the transport box to capture the top profile image of the corresponding box top cover and the surface image of the corresponding side surface of the transport box;

[0055] The transport control console is connected with the video collector to analyze the top profile image and the surface image of the side surface to determine the stability of the transport box and issue a corresponding warning according to the determination result;

[0056] The external analyzer is arranged on the transport device and is connected with the transport control console to capture the target asparagus path of the transport box during transportation;

[0057] The target asparagus path is the path of the transport box from the target asparagus position to the predetermined transport position.

[0058] Referring to Figure 2 The steps of the asparagus preservation monitoring system based on machine vision of the embodiment of the present application are shown in the figure, and the steps of the asparagus preservation monitoring system based on machine vision of the present application include:

[0059] In step S100, the target asparagus path of the transport box during transportation is obtained, the characteristic path point is determined according to the curvature of the target asparagus path, and the characteristic path segment is determined based on the characteristic path point;

[0060] The target asparagus path is the path of the transport box from the target asparagus position to the predetermined transport position.

[0061] In step S200, the video stream of the transport box moving along the characteristic path segment is collected, the top profile image of the transport box in the video stream is captured at a preset time interval, the profile size difference is determined based on each top profile image, and it is determined whether the transport box is a shaking performance characteristic transport box;

[0062] In step S300, in response to the transport box being a shaking performance characteristic transport box, the surface images of the two mutually opposite side surfaces of the shaking performance characteristic transport box are captured based on the video stream, and the dark area area of each surface image is recorded for the first time;

[0063] In step S400, in response to the shaking performance characteristic transport box moving to the predetermined transport position along the target asparagus path, the dark area area of the two side surfaces is recorded again, and whether the shaking performance characteristic transport box has a cargo scattering risk is determined according to the change of the dark area area of each surface image;

[0064] In step S500, the predetermined transfer position of the transfer box with the shaking performance feature of the cargo spill risk is determined as the spill early warning position for early warning prompt, and the target asparagus path of a plurality of transfer boxes with the shaking performance feature of the cargo spill risk is obtained, the spill negative incentive path segment is determined based on the occurrence frequency of the feature path segment in the plurality of target asparagus paths, and the curvature of the spill negative incentive path segment is adjusted according to the change of the dark area.

[0065] It can be understood that the spill negative incentive path segment is the feature path segment with the highest occurrence frequency.

[0066] Specifically, those skilled in the art can understand that asparagus transfer needs to pre-plan transfer positions for different types of asparagus, establish the association between various types of asparagus and transfer positions, and facilitate the generation of paths from target asparagus positions to predetermined transfer positions. This is prior art and will not be repeated here.

[0067] Specifically, those skilled in the art can understand that a monitoring camera can be installed around the target asparagus path of the transfer box to take real-time photos of the movement of the transfer box and obtain high-quality video streams. Video frame extraction technology can be used to capture the top profile image of the transfer box in the video stream at a preset time interval. Video stream image data acquisition technology is widely used in the fields of security, computer vision, and automation control, and will not be repeated here.

[0068] Specifically, the preset time interval for capturing the top profile image of the transfer box in the video stream can be adjusted according to the detection and evaluation accuracy requirements of the transfer box shaking and cargo spill risk. The higher the detection and evaluation accuracy requirements, the shorter the preset time interval. Preferably, the preset time interval can be in the range of [1, 3] with an interval unit of s.

[0069] Specifically, the process of determining the feature path segment based on the feature path point includes:

[0070] A rectangular coordinate system is established on the horizontal ground to determine the coordinates of a plurality of path points on the target asparagus path in the rectangular coordinate system;

[0071] Numerical differentiation is used to calculate the curvature of each path point;

[0072] The path point corresponding to the maximum curvature is determined as the feature path point;

[0073] The path segment with a preset length centered on the feature path point is determined as the feature path segment.

[0074] Specifically, the preset length of the feature path segment can be adjusted by those skilled in the art according to the requirement of detection and evaluation accuracy of the risk of the transfer box shaking and the cargo scattering. The higher the requirement of detection and evaluation accuracy is, the greater the preset length is. Preferably, the preset length can be in the range of [3, 5] with the unit of m.

[0075] Specifically, the present application determines the feature path point by the curvature of the target asparagus path, and determines the feature path segment according to the feature path point. Those skilled in the art can understand that the greater the curvature of the transfer box moving path is, the more likely the transfer box has a tendency to shake due to the larger centrifugal force. When the transfer box is transported and transported, the higher the stacking height is, and the higher the center of gravity of the transfer box is, so that the centrifugal force is more likely to affect the shaking of the transfer box. Therefore, the greater the curvature is, the worse the structural stability of the transfer box is. The present application calculates the curvature of each path point by numerical differentiation method, determines the path point corresponding to the maximum curvature as the feature path point, and determines the feature path segment according to the feature path point. Therefore, the detection and evaluation accuracy of the shaking state of the transfer box in the handling process is improved.

[0076] Specifically, the process of determining the contour size difference includes:

[0077] determining the contour maximum width value W based on each top contour image max and the contour maximum length value L max ;

[0078] calculating the contour maximum width value standard deviation σ max based on the contour maximum width value W in each top contour image W , and calculating the contour maximum length value standard deviation σ max based on the contour maximum length value L in each top contour image L ;

[0079] calculating the contour size difference representation coefficient P based on the contour maximum width value standard deviation σ W and the contour maximum length value standard deviation σ L ;

[0080] wherein the contour size difference representation coefficient P is positively correlated with the contour maximum width value standard deviation σ W , and the contour size difference representation coefficient is positively correlated with the contour maximum length value standard deviation σ L .

[0081] Specifically, the contour size difference representation coefficient P is the product of the contour maximum width value standard deviation influence factor λ and the contour maximum width value standard deviation σ W and the product of the contour maximum width value standard deviation influence factor β and the contour maximum length value standard deviation σL the sum of the product results.

[0082] Wherein, λ is the profile maximum width value standard deviation influence factor, β is the profile maximum width value standard deviation influence factor, it can be understood that the profile maximum width value standard deviation influence factor λ and the profile maximum width value standard deviation influence factor β can be selected according to the influence degree of the profile maximum width value standard deviation and the profile maximum length value standard deviation on the calculation result, λ+β=1, provide a kind of λ and β Value, λ=0.5, β=0.5.

[0083] Specifically, please refer to Figure 3 It is the logic flow chart of judging whether the transfer box is the shaking performance characteristic transfer box, the process of judging whether the transfer box is the shaking performance characteristic transfer box includes:

[0084] The profile size difference representation coefficient P is compared with the preset profile size difference representation coefficient comparison value P0;

[0085] If the profile size difference representation coefficient P is greater than the profile size difference representation coefficient comparison value P0, then the transfer box is judged as the shaking performance characteristic transfer box;

[0086] If the profile size difference representation coefficient P is less than or equal to the profile size difference representation coefficient comparison value P0, then the transfer box is not the shaking performance characteristic transfer box.

[0087] Specifically, the preset profile size difference representation coefficient comparison value P0 can be pre-tested and calculated by those skilled in the art, and the average value of the profile size difference representation coefficient of a plurality of transfer boxes moving on the target asparagus path under the same curvature condition is calculated, and the average value is determined as the profile size difference representation coefficient comparison value P0, preferably, the value range of the profile size difference representation coefficient comparison value P0 can be [8, 25], and the interval unit is cm.

[0088] Specifically, the profile size difference is determined to judge whether the transfer box is the shaking performance characteristic transfer box, and those skilled in the art can understand that the centrifugal force received by the transfer box in the characteristic path segment with large curvature can cause the overall structure of the transfer box to be distorted and deformed, and lose the original structural stability, and the top profile of the collected transfer box directly reflects the unstable condition of the internal structure of the transfer box, the more obvious the length and width size change of the top profile of the transfer box, the greater the influence on the internal structure stability of the transfer box, and the worse the stability of the transfer box, the profile size difference is determined to judge whether the transfer box is the shaking performance characteristic transfer box, and then, the shaking degree of the transfer box is directly quantified, and the detection and evaluation accuracy of the shaking and cargo scattering risk of the transfer box is improved.

[0089] Specifically, refer to Figure 4 As shown in the figure, it is a step diagram for recording the dark area of each surface image according to an embodiment of the present application, and the process of recording the dark area of each surface image comprises:

[0090] Step S301, denoising the captured surface image;

[0091] Step S302, determining the pixel points with pixel brightness lower than the preset pixel brightness threshold as dark pixels;

[0092] Step S303, determining the region composed of dark pixels as a dark region by an edge detection algorithm, and calculating the dark region area by a pixel counting method.

[0093] Specifically, the captured surface image can be denoised by using methods such as median filtering and Gaussian filtering, which are prior art and will not be described here.

[0094] Specifically, the preset pixel brightness threshold can be set according to different types of asparagus. For asparagus with darker color, the preset pixel brightness threshold can be set relatively low to more accurately identify dark pixels. For asparagus with lighter color, the threshold needs to be set relatively high to avoid misjudging normal light color regions as dark regions. For example, for purple asparagus with darker color, the pixel brightness value of the color region is 50-100, and the preset pixel brightness threshold can be set to 40. For white asparagus with lighter color, the pixel brightness value of the color region is 150-220, and the preset pixel brightness threshold can be set to 130.

[0095] Specifically, the region composed of dark pixels can be determined as a dark region by using an edge detection algorithm with the function of identifying the edge of the region according to the pixel points. The area of the dark region is calculated by counting the extracted dark pixels. Those skilled in the art can understand that edge detection algorithms and pixel counting methods are widely used in image extraction and image area measurement in the fields of medicine and agriculture, which are prior art and will not be described here.

[0096] Specifically, the process of determining the change of the dark area of each surface image comprises:

[0097] The surface images of the mutually opposite two sides of the shaking performance feature transfer box are determined as the first surface image and the second surface image;

[0098] A difference value between the dark region area of the first surface image recorded initially and the dark region area of the first surface image recorded again is determined as a first dark region area difference value S1; a difference value between the dark region area of the second surface image recorded initially and the dark region area of the second surface image recorded again is determined as a second dark region area difference value S2.

[0099] Specifically, referring to Figure 5 As shown in the figure, it is a logic flow chart for determining whether the shaking performance characteristic transfer box has the risk of cargo scattering, and the process of determining whether the shaking performance characteristic transfer box has the risk of cargo scattering comprises:

[0100] If the first dark region area difference value S1 and the second dark region area difference value S2 are opposite in sign, it is determined that the shaking performance characteristic transfer box has the risk of cargo scattering;

[0101] If the first dark region area difference value S1 and the second dark region area difference value S2 are the same in sign, it is determined that the shaking performance characteristic transfer box does not have the risk of cargo scattering.

[0102] Specifically, the present application determines whether the shaking performance characteristic transfer box has the risk of cargo scattering through the change of the dark region area of each surface image. It can be understood that the accumulated transfer boxes have accumulation voids due to the difference in asparagus shape. The image collected on the side of the transfer box will have a local dark region due to the accumulation void. The present application calculates the difference value between the dark region area in the side image of the transfer box during the target asparagus process and the dark region area in the side image of the transfer box after the target asparagus is completed, which represents the change of the accumulation void of the transfer box during the target asparagus process. If the size of the accumulation void of the two opposite sides of the transfer box changes in opposite directions, it means that the greater the change of the internal structure of the originally stable transfer box, the more likely it is to cause the transfer box to scatter, thereby improving the detection and evaluation accuracy of the shaking and cargo scattering risk of the transfer box.

[0103] Specifically, the scattering early warning position is the coordinate position of the predetermined transfer position of the shaking performance characteristic transfer box with the risk of cargo scattering in the rectangular coordinate system.

[0104] Specifically, the process of determining the scattering negative incentive path segment comprises:

[0105] Obtaining the characteristic path segment corresponding to the target asparagus path of a plurality of shaking performance characteristic transfer boxes with the risk of cargo scattering, and determining the characteristic path segment with the highest occurrence frequency as the scattering negative incentive path segment.

[0106] Specifically, the process of adjusting the curvature of the scattering negative incentive path segment comprises:

[0107] Calculating the absolute value S of the first dark region area difference valuea the absolute value S of the area difference of the second dark region b According to the average value S' of the area difference of the dark region, the curvature of the scattering negative incentive path segment is adjusted, and the curvature is negatively correlated with the average value, wherein S'=(S a +S b ) / 2.

[0108] Specifically, the curvature of the scattering negative incentive path segment is adjusted according to the change of the dark region area. As can be understood by those skilled in the art, the greater the difference value between the dark region area in the side image of the target asparagus transport box and the dark region area in the side image of the completed target asparagus transport box, the greater the shaking degree of the transport box, the greater the influence of the transport box on the feature path segment, and the curvature of the feature path segment needs to be adjusted smaller to improve the stability of the transport box, thereby realizing the timely adjustment of the target asparagus path.

[0109] So far, the technical solutions of the present application have been described in combination with the preferred embodiments shown in the drawings, but those skilled in the art can easily understand that the protection scope of the present application is obviously not limited to these specific embodiments. Those skilled in the art can make equivalent changes or replacements to the related technical features without departing from the principles of the present application, and the technical solutions after the changes or replacements will fall within the protection scope of the present application.

[0110] The above description is only the preferred embodiments of the present application and is not used to limit the present application; for those skilled in the art, the present application can have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A machine vision-based asparagus preservation monitoring system, characterized in that, It is mounted on transfer equipment, including: The transshipment box is designed as a modular box with an open front, and each part of the box is equipped with a corresponding sealing strip. The top of the box is covered with a cloth, which covers the transfer box and is sealed to each other with the sealing strips of the transfer box; A video capture device is installed above the transport box to capture the top outline image of the corresponding box top cover and the surface image of the corresponding side of the transport box. The transfer control console is connected to a video capture device to analyze the top outline image and the side surface image to determine the stability of the transfer box and issue corresponding warnings based on the determination results. An external analyzer, installed on the transfer equipment and connected to the transfer control console, is used to capture images of the target asparagus path being transferred by the transfer box. The target asparagus path is the route taken by the transfer container from the target asparagus location to the predetermined transfer location.

2. The machine vision-based asparagus preservation monitoring system according to claim 1, characterized in that, The transfer control station determines characteristic path points based on the curvature of the target asparagus path, and determines characteristic path segments based on the characteristic path points; The video acquisition device acquires a video stream of the transport box moving along the characteristic path segment, and captures the top outline image of the transport box in the video stream at preset time intervals. The transport control console determines the outline size difference based on each top outline image to determine whether the transport box is a transport box with shaking characteristics. The transfer control console, in response to the transfer box exhibiting shaking characteristics, captures surface images of two opposing sides of the transfer box based on the video stream, and initially records the area of ​​the dark region in each surface image, as well as... In response to the shaking performance characteristic transfer box moving along the target asparagus path to the predetermined transfer position, the dark area area of ​​the surface images on the two sides is recorded again, and the risk of cargo spillage is determined based on the change in the dark area area of ​​each surface image.

3. The machine vision-based asparagus preservation monitoring system according to claim 2, characterized in that, The transfer control console determines the predetermined transfer location of the transfer box with swaying characteristics that poses a risk of cargo scattering as a scattering warning location based on the target asparagus path and the risk of cargo scattering. It also acquires the target asparagus paths of several transfer boxes with swaying characteristics that pose a risk of cargo scattering, determines the scattering negative excitation path segment based on the frequency of occurrence of characteristic path segments in the target asparagus paths, and adjusts the curvature of the scattering negative excitation path segment according to the change in the area of ​​the dark region.

4. The machine vision-based asparagus preservation monitoring system according to claim 3, characterized in that, When determining the characteristic path segment, the external analyzer establishes a rectangular coordinate system on the horizontal ground and determines the coordinates of several path points on the target asparagus path in the rectangular coordinate system. The curvature of each path point is calculated using numerical differentiation. The path point corresponding to the maximum curvature is determined as the feature path point; A path segment of a preset length centered on the feature path point is defined as the feature path segment.

5. The machine vision-based asparagus preservation monitoring system according to claim 4, characterized in that, When the transfer control console determines the difference in contour dimensions, it determines the maximum width value and the maximum length value of the contour based on each top contour image. Calculate the standard deviation of the maximum width value of the contour based on the maximum width value of the contour in each top contour image, and calculate the standard deviation of the maximum length value of the contour based on the maximum length value of the contour in each top contour image; Calculate the contour size difference characterization coefficient based on the standard deviation of the maximum width value and the standard deviation of the maximum length value of the contour; The calculated profile size difference characterization coefficient is positively correlated with the standard deviation of the maximum profile width value, and the calculated profile size difference characterization coefficient is positively correlated with the standard deviation of the maximum profile length value.

6. The machine vision-based asparagus preservation monitoring system according to claim 5, characterized in that, When the transfer control console determines whether the transfer box is a transfer box with shaking characteristics, it compares the calculated contour size difference characterization coefficient with the preset contour size difference characterization coefficient comparison value. If the contour size difference characterization coefficient is greater than the contour size difference characterization coefficient comparison value, then the transport box is determined to be a transport box with shaking performance characteristics.

7. The machine vision-based asparagus preservation monitoring system according to claim 2, characterized in that, The process of recording the area of ​​the dark region in each surface image at the transfer control station includes: Denoising the captured surface image; Pixels whose brightness is lower than a preset brightness threshold are identified as dark pixels. The dark region is defined by an edge detection algorithm, and the area of ​​the dark region is calculated by a pixel counting method. The surface images of two opposing sides of the swaying performance characteristic transfer box are identified as the first surface image and the second surface image; The difference between the dark area area of ​​the first surface image recorded initially and the dark area area of ​​the first surface image recorded again is determined as the first dark area area difference value, and the difference between the dark area area of ​​the second surface image recorded initially and the dark area area of ​​the second surface image recorded again is determined as the second dark area area difference value.

8. The machine vision-based asparagus preservation monitoring system according to any one of claims 6 or 7, characterized in that, When determining whether a transshipment container exhibiting shaking characteristics poses a risk of cargo scattering, the central control station of the transfer station makes a judgment. If the area difference value of the first dark area is opposite to the area difference value of the second dark area, then the transshipment box exhibiting the shaking behavior is deemed to have a risk of cargo scattering.

9. The machine vision-based asparagus preservation monitoring system according to claim 8, characterized in that, When the transfer control console determines the negative excitation path segment for scattering, it acquires the characteristic path segments corresponding to the target asparagus path of several transfer boxes with shaking performance characteristics that have the risk of cargo scattering, and determines the characteristic path segment with the highest frequency of occurrence as the negative excitation path segment for scattering.

10. The machine vision-based asparagus preservation monitoring system according to claim 9, characterized in that, The process by which the transfer control console adjusts the curvature of the scattered negative excitation path segment includes: Calculate the average of the absolute values ​​of the area difference between the first dark region and the second dark region, and adjust the curvature of the scattered negative excitation path segment based on the average value, wherein the curvature is negatively correlated with the average value.

Images