Visual tracking system and visual tracking method for agricultural products

Abstract

The invention provides a visual tracking system and method for agricultural products, and the method comprises the steps: extracting the basic tracking information of the agricultural products from a tracking information label through analyzing the tracking information label on the outer package of the agricultural products; the basic tracking information comprises a unique identity label of the agricultural product and a first visual gradient feature, and judging whether the first visual gradient feature in the tracking information label is matched with the picture of the agricultural product; when the first visual gradient feature in the tracking information label is matched with the picture of the agricultural product, acquiring complete tracking information of the agricultural product from an agricultural product tracking server based on the unique identity label of the agricultural product, and the complete tracking information of the agricultural products is visually presented in combination with the pictures of the agricultural products, so that the loopholes of the existing anti-counterfeiting means can be effectively filled, and the agricultural products which are not strictly detected are prevented from flowing into the market.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural product tracking, and particularly relates to a visual tracking system and a visual tracking method for agricultural products. Background Art

[0002] Food safety issues have always been important livelihood issues that people have been extremely concerned about. Due to the large number of batches, large quantity, and complex distribution channels of agricultural products, when quality problems occur in any batch of agricultural products, without effective tracking, it will cause extremely huge losses to agricultural products from the same source. Therefore, in food safety issues, the quality problems of agricultural products, especially those caused by pollution or excessive pesticide residues, have always been difficult regulatory problems. To solve the problem of difficult supervision of agricultural products, people have developed technologies such as using two-dimensional codes and blockchains for food safety tracking, recording the associated information of agricultural products in all links of production, processing, transportation, and sales, including information such as the place of origin, planting method, usage of pesticides and fertilizers, production date, processing method, inspection report, transportation method, and associated entities in each link such as producers, processors, carriers, and distributors. This makes it possible to trace the complete information of agricultural products when quality problems occur, improves the response decision-making efficiency, and reduces unnecessary losses.

[0003] In existing agricultural product tracking solutions, the most common is to print the tracking information of agricultural products in the form of two-dimensional codes on the outer packaging of agricultural products. Since packaging and printing technologies are not complex and the production cost is relatively low, some unscrupulous merchants may forge or replace the outer packaging to achieve the purpose of passing off inferior goods as superior ones. In particular, directly replacing the agricultural products in the existing packaging or repackaging with recycled packaging makes various anti-counterfeiting technologies unable to achieve the anti-counterfeiting effect, resulting in some agricultural products that have not been strictly inspected flowing into the market. Summary of the Invention

[0004] Based on the above problems, the present invention proposes a visual tracking system and a visual tracking method for agricultural products, which can effectively fill the loopholes in existing anti-counterfeiting means and prevent agricultural products that have not been strictly inspected from flowing into the market.

[0005] In view of this, the first aspect of the present invention proposes a visual tracking system for agricultural products, including a camera device for taking pictures of agricultural products, a barcode recognition device for capturing and identifying two-dimensional barcodes on the outer packaging of agricultural products, an agricultural product tracking server for storing the complete tracking information of agricultural products, and a visual tracking client connected to the camera device, the barcode recognition device, and the agricultural product tracking server. The visual tracking client is configured to: Obtain a photo of the agricultural product and a tracking information label, where the tracking information label is a text or graphic label printed on the outer package of the agricultural product and storing the tracking information of the agricultural product; Analyze the tracking information label to extract the basic tracking information of the agricultural product from the tracking information label, where the basic tracking information includes the unique identity identifier of the agricultural product and the first visualization gradient feature; Determine whether the first visualization gradient feature in the tracking information label matches the photo of the agricultural product; When the first visualization gradient feature in the tracking information label matches the photo of the agricultural product, obtain the complete tracking information of the agricultural product from the agricultural product tracking server based on the unique identity identifier of the agricultural product; Visually present the complete tracking information of the agricultural product in combination with the photo of the agricultural product.

[0006] A second aspect of the present invention proposes a method for visual tracking of agricultural products, including: Obtain a photo of the agricultural product and a tracking information label, where the tracking information label is a text or graphic label printed on the outer package of the agricultural product and storing the tracking information of the agricultural product; Analyze the tracking information label to extract the basic tracking information of the agricultural product from the tracking information label, where the basic tracking information includes the unique identity identifier of the agricultural product and the first visualization gradient feature; Determine whether the first visualization gradient feature in the tracking information label matches the photo of the agricultural product; When the first visualization gradient feature in the tracking information label matches the photo of the agricultural product, obtain the complete tracking information of the agricultural product from the agricultural product tracking server based on the unique identity identifier of the agricultural product; Visually present the complete tracking information of the agricultural product in combination with the photo of the agricultural product.

[0007] Further, before the step of obtaining the photo of the agricultural product and the tracking information label, it further includes: Before packaging the agricultural product, take a photo of the agricultural product; Extract the first visualization gradient feature of the agricultural product from the photo of the agricultural product; Generate the unique identity identifier of the agricultural product; Write the unique identity identifier of the agricultural product and the first visualization gradient feature into the tracking information label of the agricultural product; Print or attach the tracking information label to the outer package of the agricultural product.

[0008] Further, the step of extracting the first visual gradient feature of the agricultural product from the photo of the agricultural product specifically includes: Configuring an identification scale for extracting key identification features; Extracting the main image of the agricultural product from the photo of the agricultural product; Performing denoising processing and smoothing processing on the main image; Finding local extreme points in the main image after denoising processing and smoothing processing based on the identification scale; Determining the local extreme points as the feature points of the main image; Extracting the first visual gradient feature of the agricultural product at the feature points on the main image.

[0009] Further, the step of extracting the first visual gradient feature of the agricultural product at the local extreme points specifically includes: Configuring a gradient statistical scale for counting the first visual gradient features of the feature points, where the first visual gradient features include the gradient feature values and gradient feature directions of the feature points; Taking each feature point on the main image as the center and the gradient statistical scale as the radius to determine the gradient statistical scale range corresponding to the feature point; Calculating the gradient feature values and gradient feature directions of the feature points according to the gradient features of each pixel point within the gradient statistical scale range of the feature points; Determining the gradient feature values and gradient feature directions at each feature point on the main image as the first visual gradient features of the agricultural product.

[0010] Further, the step of calculating the gradient feature values and gradient feature directions of the feature points according to the gradient features of each pixel point within the gradient statistical scale range of the feature points specifically includes: Traversing each feature point on the main image; Determining the traversed feature point as the target feature point; Traversing the gradient amplitudes and gradient directions of each pixel within the gradient statistical scale range of the target feature point where and are respectively the horizontal coordinate and vertical coordinate of the corresponding pixel on the main image; Counting the number of pixel points within the gradient statistical scale range of the target feature point whose gradient directions fall into each angular interval where is a positive integer from 1 to ; is the number of angular intervals divided by 360 with a preset as the interval length, and satisfies: ; Calculate the cumulative gradient magnitude of each angular interval : ; where is a positive integer from 1 to , is the horizontal coordinate of the th pixel within the gradient statistical scale range of the target feature point, is the vertical coordinate of the th pixel within the gradient statistical scale range of the target feature point; Determine a positive integer between 1 and such that the cumulative gradient magnitude where the gradient direction falls within the angular interval satisfies: ; Determine the cumulative gradient magnitude of the th angular interval as the gradient eigenvalue of the target feature point; Determine the th angular interval as the gradient feature direction of the target feature point.

[0011] Furthermore, the step of determining whether the first visual gradient feature in the tracking information label matches the photo of the agricultural product specifically includes: Extract the second visual gradient feature of the agricultural product from the photo of the agricultural product; Compare the first visual gradient feature with the second visual gradient feature to determine whether the second visual gradient feature is the same as the first visual gradient feature; When the second visual gradient feature is the same as the first visual gradient feature, determine that the first visual gradient feature in the tracking information label matches the photo of the agricultural product.

[0012] Furthermore, the step of comparing the first visual gradient feature with the second visual gradient feature to determine whether the second visual gradient feature is the same as the first visual gradient feature specifically includes: Obtain the number of feature points of the first visual gradient feature and the number of feature points of the second visual gradient feature ; Calculate the number of feature points of the first visual gradient feature and the number of feature points of the second visual gradient feature The difference in the number of feature points: ; When the difference in the number of feature points is less than or equal to a preset deviation threshold, establish the correspondence relationship of feature points between the first visual gradient feature and the second visual gradient feature; Compare the gradient feature directions between the feature points with corresponding relationships in the first visual gradient feature and the second visual gradient feature; When the gradient directions of the feature points with corresponding relationships in the first visual gradient feature and the second visual gradient feature are in the same angular interval, calculate the deviation of the gradient feature values between the feature points with corresponding relationships in the first visual gradient feature and the second visual gradient feature; When the deviation of the gradient feature values between the feature points with corresponding relationships in the first visual gradient feature and the second visual gradient feature is less than or equal to a preset deviation threshold, determine that the second visual gradient feature is the same as the first visual gradient feature.

[0013] Further, the first visual gradient feature and the second visual gradient feature include the pixel coordinates of each feature point. The steps of establishing the correspondence relationship of feature points between the first visual gradient feature and the second visual gradient feature specifically include: Construct a first feature point topology based on the pixel coordinates of the feature points included in the first visual gradient feature and a second feature point topology based on the pixel coordinates of the feature points included in the second visual gradient feature; Compare the topological shapes of the first feature point topology and the second feature point topology; Judge whether the topological shapes of the first feature point topology and the second feature point topology are the same; When the topological shapes of the first feature point topology and the second feature point topology are different, judge whether the topological shapes of the first feature point topology and the second feature point topology satisfy that one is a subset of the other; When the topological shapes of the first feature point topology and the second feature point topology are the same, or the topological shapes of the first feature point topology and the second feature point topology satisfy that one is a subset of the other, determine that the feature points of the first visual gradient feature and the second visual gradient feature have a corresponding relationship; Establish a correspondence between the feature points at the corresponding positions in the first feature point topology and the second feature point topology.

[0014] Further, the steps of visually presenting the complete tracking information of the agricultural product in combination with the photo of the agricultural product specifically include: Obtain the node tracking information of the agricultural product at each tracking node in the complete tracking information of the agricultural product, where the tracking node information includes the photo of the agricultural product reported by the corresponding tracking node and its visual gradient feature; Present the photos of the agricultural product at each tracking node in chronological order; Fuse and present the visual gradient features of the agricultural product at each tracking node on the photo of the corresponding tracking node.

[0015] The present invention proposes a visual tracking system and a visual tracking method for agricultural products. By parsing the tracking information label on the outer package of the agricultural product to extract the basic tracking information of the agricultural product from the tracking information label, the basic tracking information includes the unique identity identifier of the agricultural product and the first visual gradient feature, determine whether the first visual gradient feature in the tracking information label matches the photo of the agricultural product. When the first visual gradient feature in the tracking information label matches the photo of the agricultural product, obtain the complete tracking information of the agricultural product from the agricultural product tracking server based on the unique identity identifier of the agricultural product, and visually present the complete tracking information of the agricultural product in combination with the photo of the agricultural product, which can effectively fill the loopholes in existing anti-counterfeiting means and prevent agricultural products that have not been strictly tested from flowing into the market. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 is a schematic diagram of a visual tracking system for agricultural products provided by an embodiment of the present invention; Figure 2 is a flowchart of a visual tracking method for agricultural products provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] In order to more clearly understand the above objects, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, without conflict, the embodiments of the present application and the features in the embodiments can be combined with each other.

[0018] Many specific details are set forth in the following description in order to fully understand the present invention. However, the present invention may be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited by the specific embodiments disclosed below.

[0019] In the description of the present invention, the term "a plurality of" refers to two or more, unless otherwise clearly defined. The orientation or positional relationship indicated by terms such as "upper", "lower", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention. Terms such as "connection", "installation", "fixation", etc. should all be understood in a broad sense. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; it can be directly connected or indirectly connected through an intermediate medium. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances. In addition, terms such as "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first", "second", etc. may explicitly or implicitly include one or more of such features. In the description of the present invention, unless otherwise stated, the meaning of "a plurality of" is two or more.

[0020] In the description of this specification, the description of terms such as "an embodiment", "some embodiments", "specific embodiments", etc. means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic expressions of the above terms do not necessarily refer to the same embodiment or instance. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0021] Next, a visualization tracking system and a visualization tracking method for agricultural products provided according to some embodiments of the present invention will be described with reference to the accompanying drawings.

[0022] As Figure 1 shown, a first aspect of the present invention provides a visualization tracking system for agricultural products, including a camera device for taking photos of agricultural products, a barcode recognition device for capturing and identifying two-dimensional barcodes on the outer packaging of agricultural products, an agricultural product tracking server for storing complete tracking information of agricultural products, and a visualization tracking client connected to the camera device, the barcode recognition device, and the agricultural product tracking server. The visualization tracking client is configured to: Obtain photos of agricultural products and tracking information labels, where the tracking information labels are text or graphic labels printed on the outer packaging of the agricultural products and storing the tracking information of the agricultural products; Parse the tracking information label to extract the basic tracking information of the agricultural product from the tracking information label, where the basic tracking information includes the unique identity identifier of the agricultural product and the first visual gradient feature; Determine whether the first visual gradient feature in the tracking information label matches the photo of the agricultural product; When the first visual gradient feature in the tracking information label matches the photo of the agricultural product, obtain the complete tracking information of the agricultural product from the agricultural product tracking server based on the unique identity identifier of the agricultural product; Visually present the complete tracking information of the agricultural product in combination with the photo of the agricultural product.

[0023] Specifically, the photo of the agricultural product refers to the photo of the agricultural product after removing the packaging, which can be the photo of the agricultural product purchased by the consumer or the photo taken after removing the packaging of the agricultural product sampled by the inspector. In some embodiments of the present invention, the photo of the agricultural product can also be the photo taken by the producer or processor before packaging the agricultural product. The producer or processor prints the taken photo on the packaging of the agricultural product. This photo has the same appearance and form as the agricultural product to be packaged, so that before purchasing the agricultural product, the consumer can intuitively see whether the photo is consistent with the agricultural product in the packaging, and accordingly decide whether to purchase.

[0024] Further, the photo of the agricultural product can be a single-sided photo of the agricultural product with key identification features, or a multi-sided photo of the agricultural product. The single-sided photo refers to a photo taken of one orientation of the agricultural product, and the multi-sided photo refers to multiple photos taken of multiple angles of the agricultural product.

[0025] The first visual gradient feature is a key identification feature on the appearance of the agricultural product. More specifically, the first visual gradient feature is an identification feature that reflects the gradient information of some specific feature points on the appearance of the agricultural product.

[0026] Further, before the steps of obtaining the photo and the tracking information label of the agricultural product, the visual tracking client is configured to: Take a photo of the agricultural product before packaging the agricultural product; Extract the first visual gradient feature of the agricultural product from the photo of the agricultural product; Generate the unique identity identifier of the agricultural product; Write the unique identity identifier and the first visual gradient feature of the agricultural product into the tracking information label of the agricultural product; Print or attach the tracking information label to the outer packaging of the agricultural product.

[0027] As described above, the tracking information label can be a graphic label or a label composed of a combination of text and graphics. For example, the unique identity identifier of the agricultural product included in the tracking information label can be a serial number composed of numbers, letters, and / or special symbols, etc., presented in the form of a text label on the outer packaging of the agricultural product. The first visual gradient feature is gradient data of key recognition features extracted from a photo of the agricultural product, which is encoded and printed on the outer packaging of the agricultural product in the form of a graphic label such as a QR code. It should be noted that the first visual gradient feature in the tracking information label is gradient data extracted from a photo taken before the agricultural product is packaged.

[0028] Further, in the step of extracting the first visual gradient feature of the agricultural product from the photo of the agricultural product, the visual tracking client is configured to: Configure an identification scale for extracting key recognition features; Extract the main image of the agricultural product from the photo of the agricultural product; Perform denoising processing and smoothing processing on the main image; Find local extreme points in the main image after performing denoising processing and smoothing processing based on the identification scale; Determine the local extreme points as the feature points of the main image; Extract the first visual gradient feature of the agricultural product at the feature points on the main image.

[0029] Specifically, the main image refers to an image that only contains the agricultural product itself and does not contain any external environmental background. The purpose of performing denoising processing and smoothing processing is to eliminate noise points and singular points on the main image and avoid the extracted extreme points being merely noise or unimportant textures on the main image. Preferably, after the step of performing denoising processing and smoothing processing on the main image, gray-scale processing is further included to remove the color information on the main image.

[0030] Further, the identification scale is the pixel width of the extreme point search window for finding local extreme points. The extreme point search window is a rectangular window. In the step of finding local extreme points in the main image after performing denoising processing and smoothing processing based on the identification scale, the visual tracking client is configured to: Configure a local extreme point search step for moving the extreme point search window, and the local extreme point search step is less than the pixel width of the extreme point search window; Fill the main image with pixels of a preset color to construct the main image as a rectangular image; Starting from the upper left corner of the rectangular image, gradually move the extreme point search window with the step size of the local extreme point search as the moving step size; After each movement, determine the extreme points within the extreme point search window as the local extreme points.

[0031] Furthermore, in the step of filling the main image with pixels of a preset color to construct the main image into a rectangular image, the preset color can be any gray scale value configured in advance. Preferably, the preset color is black or white.

[0032] Furthermore, in the step of extracting the first visual gradient feature of the agricultural product at the local extreme point, the visual tracking client is configured to: Configure a gradient statistical scale for statistically calculating the first visual gradient feature of the feature points, where the first visual gradient feature includes the gradient feature value and the gradient feature direction of the feature points; Taking each feature point on the main image as the center and the gradient statistical scale as the radius, determine the gradient statistical scale range corresponding to the feature point; Calculate the gradient feature value and the gradient feature direction of the feature point according to the gradient features of each pixel point within the gradient statistical scale range of the feature point; Determine the gradient feature values and the gradient feature directions at each feature point on the main image as the first visual gradient feature of the agricultural product.

[0033] Specifically, the first visual gradient feature of the agricultural product is composed of the first visual gradient features of each feature point on the main image. In the technical solution of the above embodiment, the first visual gradient feature of each feature point on the main image reflects the comprehensive gradient information of the pixel points within the gradient statistical scale range with it as the center, that is, the gradient feature value and the gradient feature direction in the first visual gradient feature of each feature point are calculated from the gradient information of the pixel points within the gradient statistical scale range with this feature point as the center.

[0034] Furthermore, in the step of calculating the gradient feature value and the gradient feature direction of the feature point according to the gradient features of each pixel point within the gradient statistical scale range of the feature point, the visual tracking client is configured to: Traverse each feature point on the main image; Determine the traversed feature point as the target feature point; Traverse the gradient amplitude of each pixel in the gradient statistical scale range of the target feature point and the gradient direction , where and are the horizontal coordinate and the vertical coordinate of the corresponding pixel on the main image, respectively; Count the number of pixel points within the gradient statistical scale range of the target feature point whose gradient directions fall into each angular interval where , where is a positive integer between 1 and , is the number of angular intervals divided from 360 degrees with a preset as the interval length, and satisfies: ; Calculate the cumulative gradient amplitude of each angular interval : ; where is a positive integer between 1 and , is the horizontal coordinate of the th pixel within the gradient statistical scale range of the target feature point, is the vertical coordinate of the th pixel within the gradient statistical scale range of the target feature point; Determine a positive integer between 1 and such that the cumulative gradient amplitude falling into the angular interval satisfies: ; Set the cumulative gradient amplitude of the th angular interval as the gradient eigenvalue of the target feature point; Set the th angular interval as the gradient feature direction of the target feature point.

[0035] In the technical solution of the above embodiment, the lower bound is denoted as , and the upper bound is denoted as , for any value between 1 and (including ) satisfies: , and .

[0036] Since is one of the positive integers between 1 and , therefore One of the values among. Similarly, is one of the values among. Similarly, is one of the angular intervals among, whose lower bound is expressed as and the upper bound is expressed as .

[0037] Furthermore, in the step of traversing the gradient magnitude and gradient direction of each pixel in the gradient statistical scale range of the target feature points, the visualization tracking client is configured to: represent the pixel value at each pixel in the main body image as a function; calculate the partial derivative of the main body image in the direction of at the pixel point : ; calculate the partial derivative of the main body image in the direction of at the pixel point : ; calculate the gradient magnitude of the main body image at the pixel point : ; calculate the gradient direction of the main body image at the pixel point : .

[0038] Furthermore, in the step of determining whether the first visualization gradient feature in the tracking information label matches the photo of the agricultural product, the visualization tracking client is configured to: extract the second visualization gradient feature of the agricultural product from the photo of the agricultural product; compare the first visualization gradient feature with the second visualization gradient feature to determine whether the second visualization gradient feature is the same as the first visualization gradient feature; when the second visualization gradient feature is the same as the first visualization gradient feature, determine that the first visualization gradient feature in the tracking information label matches the photo of the agricultural product.

[0039] In the technical solution of the above embodiment, the first visual gradient feature is a key recognition feature extracted from a photo taken before the agricultural product is packaged, which reflects the appearance feature of the agricultural product before it is packaged and put on the market. The second visual gradient feature is a key recognition feature extracted from a photo taken after the agricultural product is packaged, during sampling inspection or after being purchased by a consumer, which reflects the appearance feature of the agricultural product after it is packaged and put on the market. Therefore, when the second visual gradient feature is the same as the first visual gradient feature, the first visual gradient feature in the tracking information label can be matched with the photo of the agricultural product, that is, it can be determined that the agricultural product inside the package has not been maliciously replaced.

[0040] Further, in the step of comparing the first visual gradient feature with the second visual gradient feature to determine whether the second visual gradient feature is the same as the first visual gradient feature, the visual tracking client is configured to: Obtain the number of feature points of the first visual gradient feature and the number of feature points of the second visual gradient feature ; Calculate the number of feature points of the first visual gradient feature and the number of feature points of the second visual gradient feature of the difference in the number of feature points: ; When the difference in the number of feature points is less than or equal to a preset deviation threshold, establish a corresponding relationship between the feature points of the first visual gradient feature and the second visual gradient feature; Compare the gradient feature directions between the feature points with corresponding relationships in the first visual gradient feature and the second visual gradient feature; When the gradient directions of the feature points with corresponding relationships in the first visual gradient feature and the second visual gradient feature are in the same angular range, calculate the deviation of the gradient feature values between the feature points with corresponding relationships in the first visual gradient feature and the second visual gradient feature; When the deviation of the gradient feature values between the feature points with corresponding relationships in the first visual gradient feature and the second visual gradient feature is less than or equal to a preset deviation threshold, determine that the second visual gradient feature is the same as the first visual gradient feature.

[0041] It should be noted that due to factors such as the shooting environment (especially the environmental brightness), the shooting device, the shooting angle, and a certain degree of change in the appearance or shape of the agricultural product itself, the second visual gradient feature and the first visual gradient feature often do not completely coincide. Therefore, when it is said in the present invention that the second visual gradient feature is the same as the first visual gradient feature, it means that the deviation between the second visual gradient feature and the first visual gradient feature is within a certain threshold range.

[0042] Further, in the step of establishing the corresponding relationship of the feature points between the first visual gradient feature and the second visual gradient feature, where the first visual gradient feature and the second visual gradient feature include the pixel coordinates of each feature point, the visual tracking client is configured to: Construct a first feature point topology based on the pixel coordinates of the feature points included in the first visual gradient feature and a second feature point topology based on the pixel coordinates of the feature points included in the second visual gradient feature; Compare the topological shapes of the first feature point topology and the second feature point topology; Determine whether the topological shapes of the first feature point topology and the second feature point topology are the same; When the topological shapes of the first feature point topology and the second feature point topology are different, determine whether the topological shapes of the first feature point topology and the second feature point topology satisfy that one is a subset of the other; When the topological shapes of the first feature point topology and the second feature point topology are the same, or when the topological shapes of the first feature point topology and the second feature point topology satisfy that one is a subset of the other, determine that the feature points of the first visual gradient feature and the second visual gradient feature have a corresponding relationship; Establish the corresponding relationship between the feature points at the corresponding positions in the first feature point topology and the second feature point topology.

[0043] The pixel coordinates of the feature points included in the first visual gradient feature and the second visual gradient feature refer to the horizontal coordinate and the vertical coordinate of the feature points on the main image.

[0044] The first feature point topology and the second feature point topology are topological graphs with the pixel points corresponding to the pixel coordinates of the feature points as nodes and the connections between the nodes as the node relationships. When the second visual gradient feature is the same as the first visual gradient feature, the first feature point topology and the second feature point topology should have the same structure, or satisfy the relationship that one is a subset of the other.

[0045] The relationship that one of the first feature point topology and the second feature point topology satisfies being a subset of the other can be that the number of nodes in the first feature point topology is less than the number of nodes in the second feature point topology, and after removing the extra nodes in the second feature point topology, it has the same structure as the first feature point topology, and vice versa.

[0046] Further, in the step of visually presenting the complete traceability information of the agricultural product in combination with the photo of the agricultural product, the visual traceability client is configured to: Obtain the node traceability information of the agricultural product at each traceability node in the complete traceability information of the agricultural product, where the traceability node information includes the photo of the agricultural product reported by the corresponding traceability node and its visual gradient feature; Present the photos of the agricultural product at each traceability node in chronological order; Fuse and present the visual gradient features of the agricultural product at each traceability node on the photo of the corresponding traceability node.

[0047] Specifically, the traceability node refers to the merchant node that reports the traceability information of the agricultural product to the agricultural product traceability server each time, which can be a producer, processor, inspection contractor, carrier or distributor, etc. The node traceability information at least includes information such as the scanning time, scanning location, photo of the agricultural product, visual gradient feature, and merchant ID.

[0048] Further, in the step of fusing and presenting the visual gradient features of the agricultural product at each traceability node on the photo of the corresponding traceability node, the visual traceability client is configured to: Respectively determine the photo of the agricultural product and the visual gradient feature in the node traceability information of the same traceability node as the target photo and the target gradient feature; Traverse each feature point in the target gradient feature to obtain the gradient feature value and gradient feature direction of the corresponding feature point; Obtain the pixel coordinates of each feature point in the target gradient feature on the target photo; Draw a circle on the target photo with the pixel coordinates as the center and the gradient feature value as the radius; Draw a feature line of the corresponding feature point on the target photo with the pixel coordinates as the starting point and the central angle of the gradient feature direction as the extension direction, and the end point of the feature line falls on the circle.

[0049] As Figure 2 shown, the second aspect of the present invention proposes a visual traceability method for agricultural products, including: Obtain a photo of the agricultural product and a tracking information label, where the tracking information label is a text or graphic label printed on the outer package of the agricultural product and storing the tracking information of the agricultural product; Analyze the tracking information label to extract the basic tracking information of the agricultural product from the tracking information label, where the basic tracking information includes the unique identity identifier of the agricultural product and the first visual gradient feature; Determine whether the first visual gradient feature in the tracking information label matches the photo of the agricultural product; When the first visual gradient feature in the tracking information label matches the photo of the agricultural product, obtain the complete tracking information of the agricultural product from the agricultural product tracking server based on the unique identity identifier of the agricultural product; Visually present the complete tracking information of the agricultural product in combination with the photo of the agricultural product.

[0050] Specifically, the photo of the agricultural product refers to the photo of the agricultural product after removing the package, which can be the photo of the agricultural product purchased by the consumer or the photo taken after sampling by the inspector after removing the package of the agricultural product. In the technical solutions of some embodiments of the present invention, the photo of the agricultural product can also be the photo taken by the producer or processor before packaging the agricultural product. The producer or processor prints the taken photo on the agricultural product package, and the photo has the same appearance and form as the agricultural product to be packaged, so that the consumer can intuitively see whether the photo is consistent with the agricultural product in the package before purchasing the agricultural product, and thus decide whether to purchase.

[0051] Further, the photo of the agricultural product can be a single-sided photo of the agricultural product with key identification features, or a multi-sided photo of the agricultural product. The single-sided photo refers to a photo taken on one orientation of the agricultural product, and the multi-sided photo refers to multiple photos taken from multiple angles of the agricultural product.

[0052] The first visual gradient feature is a key identification feature in the appearance of the agricultural product. More specifically, the first visual gradient feature is an identification feature reflecting the gradient information of some specific feature points on the appearance of the agricultural product.

[0053] Further, before the step of obtaining the photo of the agricultural product and the tracking information label, it further includes: Before packaging the agricultural product, take a photo of the agricultural product; Extract the first visual gradient feature of the agricultural product from the photo of the agricultural product; Generate the unique identity identifier of the agricultural product; Write the unique identity identifier of the agricultural product and the first visual gradient feature into the tracking information label of the agricultural product; Print or attach the tracking information label to the outer package of the agricultural product.

[0054] As mentioned above, the tracking information label can be a graphic label or a label composed of a combination of text and graphics. For example, the unique identity identifier of the agricultural product contained in the tracking information label can be a serial number composed of numbers, letters, and / or special symbols, etc., presented in the form of a text label on the outer package of the agricultural product, and the first visual gradient feature is the gradient data of the key recognition features extracted from the photo of the agricultural product, which is printed on the outer package of the agricultural product in the form of a graphic label such as a QR code after being encoded. It should be known that the first visual gradient feature in the tracking information label is the gradient data extracted from the photo taken before the agricultural product is packaged.

[0055] Further, the steps of extracting the first visual gradient feature of the agricultural product from the photo of the agricultural product specifically include: Configure the recognition scale for extracting key recognition features; Extract the main body image of the agricultural product from the photo of the agricultural product; Perform denoising processing and smoothing processing on the main body image; Find local extreme points in the main body image after performing denoising processing and smoothing processing based on the recognition scale; Determine the local extreme points as the feature points of the main body image; Extract the first visual gradient feature of the agricultural product at the feature points on the main body image.

[0056] Specifically, the main body image refers to an image that only contains the agricultural product itself and does not contain any external environmental background. The purpose of performing denoising processing and smoothing processing is to eliminate the noise points and singular points on the main body image and avoid the extracted extreme points being just the noise or unimportant textures on the main body image. Preferably, after the step of performing denoising processing and smoothing processing on the main body image, gray-scale processing is also included to remove the color information on the main body image.

[0057] Further, the recognition scale is the pixel width of the extreme point search window for finding local extreme points. The extreme point search window is a rectangular window. The steps of finding local extreme points in the main body image after performing denoising processing and smoothing processing based on the recognition scale specifically include: Configure the local extreme point search step size for moving the extreme point search window, where the local extreme point search step size is less than the pixel width of the extreme point search window; Fill the main image with pixels of a preset color to construct the main image into a rectangular image; Starting from the upper left corner of the rectangular image, move the extreme point search window step by step with the local extreme point search step size as the moving step size; After each move, determine the extreme points within the extreme point search window as the local extreme points.

[0058] Further, in the step of filling the main image with pixels of a preset color to construct the main image into a rectangular image, the preset color can be any pre-configured grayscale value. Preferably, the preset color is black or white.

[0059] Further, the step of extracting the first visual gradient feature of the agricultural product at the local extreme point specifically includes: Configure the gradient statistical scale for counting the first visual gradient feature of the feature points, where the first visual gradient feature includes the gradient feature value and gradient feature direction of the feature points; Taking each feature point on the main image as the center and the gradient statistical scale as the radius, determine the gradient statistical scale range corresponding to the feature point; Calculate the gradient feature value and gradient feature direction of the feature point according to the gradient features of each pixel point within the gradient statistical scale range of the feature point; Determine the gradient feature value and gradient feature direction at each feature point on the main image as the first visual gradient feature of the agricultural product.

[0060] Specifically, the first visual gradient feature of the agricultural product is composed of the first visual gradient features of each feature point on the main image. In the technical solution of the above embodiment, the first visual gradient feature of each feature point on the main image reflects the comprehensive gradient information of the pixel points within the gradient statistical scale range with it as the center, that is, the gradient feature value and gradient feature direction in the first visual gradient feature of each feature point are calculated from the gradient information of the pixel points within the gradient statistical scale range with this feature point as the center.

[0061] Further, the step of calculating the gradient feature value and gradient feature direction of the feature point according to the gradient features of each pixel point within the gradient statistical scale range of the feature point specifically includes: Traverse each feature point on the main image; Determine the traversed feature point as the target feature point; Traverse the gradient statistical scale range of the target feature point and the gradient amplitude of each pixel and the gradient direction ,in and are the horizontal coordinate and the vertical coordinate of the corresponding pixel on the subject image respectively; Count the gradient direction of the target feature point within the gradient statistics scale and fall into each angle interval The number of pixels ,in 1 to A positive integer between 360 degrees divided into preset is the number of angle intervals of the interval length, and satisfy: ; Calculate each angle interval The cumulative gradient amplitude of : ; in 1 to A positive integer between is the first The horizontal coordinate of the pixel, is the first The vertical coordinate of the pixel; In 1 to Determine a positive integer between , so that the gradient direction falls into the angle interval The cumulative gradient amplitude satisfy: ; The first The cumulative gradient amplitude of the angle interval Determining the gradient feature value of the target feature point; The first Angle interval The gradient feature direction of the target feature point is determined.

[0062] In the technical solution of the above-mentioned implementation mode, The lower bound of , the upper bound is expressed as , for 1 to Between (inclusive ) The value satisfies: ,and .

[0063] Since is one of the positive integers from 1 to , therefore is one of the values among. Similarly, is one of the angular intervals among, the lower bound of which is represented as , and the upper bound is represented as .

[0064] Furthermore, the steps of traversing the gradient magnitude and gradient direction of each pixel in the gradient statistical scale range of the target feature points specifically include: Represent the pixel value at each pixel in the main image as a function; Calculate the partial derivative of the main image in the direction of at the pixel point : ; Calculate the partial derivative of the main image in the direction of at the pixel point : ; Calculate the gradient magnitude of the main image at the pixel point : ; Calculate the gradient direction of the main image at the pixel point : .

[0065] Furthermore, the steps of determining whether the first visual gradient feature in the tracking information label matches the photo of the agricultural product specifically include: Extract the second visual gradient feature of the agricultural product from the photo of the agricultural product; Compare the first visual gradient feature with the second visual gradient feature to determine whether the second visual gradient feature is the same as the first visual gradient feature; When the second visual gradient feature is the same as the first visual gradient feature, determine that the first visual gradient feature in the tracking information label matches the photo of the agricultural product.

[0066] In the technical solution of the above embodiment, the first visual gradient feature is a key recognition feature extracted from a photo taken before the agricultural product is packaged, which reflects the appearance feature of the agricultural product before it is packaged and put on the market. The second visual gradient feature is a key recognition feature extracted from a photo taken after the agricultural product is packaged, during sampling inspection or after being purchased by a consumer, which reflects the appearance feature of the agricultural product after it is packaged and put on the market. Therefore, when the second visual gradient feature is the same as the first visual gradient feature, the first visual gradient feature in the tracking information label can be matched with the photo of the agricultural product, that is, it can be determined that the agricultural product inside the package has not been maliciously replaced.

[0067] Further, the step of comparing the first visual gradient feature with the second visual gradient feature to determine whether the second visual gradient feature is the same as the first visual gradient feature specifically includes: Obtain the number of feature points of the first visual gradient feature and the number of feature points of the second visual gradient feature ; Calculate the difference in the number of feature points between the number of feature points of the first visual gradient feature and the number of feature points of the second visual gradient feature : ; When the difference in the number of feature points is less than or equal to a preset deviation threshold, establish a corresponding relationship between the feature points of the first visual gradient feature and the second visual gradient feature; Compare the gradient feature directions between the feature points with corresponding relationships in the first visual gradient feature and the second visual gradient feature; When the gradient directions of the feature points with corresponding relationships in the first visual gradient feature and the second visual gradient feature are in the same angular range, calculate the deviation of the gradient feature values between the feature points with corresponding relationships in the first visual gradient feature and the second visual gradient feature; When the deviation of the gradient feature values between the feature points with corresponding relationships in the first visual gradient feature and the second visual gradient feature is less than or equal to a preset deviation threshold, determine that the second visual gradient feature is the same as the first visual gradient feature.

[0068] It should be noted that due to factors such as the shooting environment (especially the environmental brightness), the shooting device, the shooting angle, and certain changes in the appearance or shape of the agricultural product itself, the second visual gradient feature and the first visual gradient feature often will not be exactly the same. Therefore, when it is said in the present invention that the second visual gradient feature is the same as the first visual gradient feature, it means that the deviation between the second visual gradient feature and the first visual gradient feature is within a certain threshold range.

[0069] Further, the first visual gradient feature and the second visual gradient feature include the pixel coordinates of each feature point. The steps of establishing the corresponding relationship between the feature points of the first visual gradient feature and the second visual gradient feature specifically include: Constructing a first feature point topology based on the pixel coordinates of the feature points included in the first visual gradient feature and a second feature point topology based on the pixel coordinates of the feature points included in the second visual gradient feature; Comparing the topological shapes of the first feature point topology and the second feature point topology; Judging whether the topological shapes of the first feature point topology and the second feature point topology are the same; When the topological shapes of the first feature point topology and the second feature point topology are not the same, judging whether the topological shapes of the first feature point topology and the second feature point topology satisfy that one is a subset of the other; When the topological shapes of the first feature point topology and the second feature point topology are the same, or the topological shapes of the first feature point topology and the second feature point topology satisfy that one is a subset of the other, determining that the feature points of the first visual gradient feature and the second visual gradient feature have a corresponding relationship; Establishing the corresponding relationship between the feature points at the corresponding positions in the first feature point topology and the second feature point topology.

[0070] The pixel coordinates of the feature points included in the first visual gradient feature and the second visual gradient feature refer to the horizontal coordinate and the vertical coordinate of the feature point on the main image.

[0071] The first feature point topology and the second feature point topology are topological graphs with the pixel points corresponding to the pixel coordinates of the feature points as nodes and the connections between the nodes as the node relationships. When the second visual gradient feature is the same as the first visual gradient feature, the first feature point topology and the second feature point topology should have the same structure, or satisfy the relationship that one is a subset of the other.

[0072] The relationship that one of the first feature point topologies and the second feature point topologies satisfies being a subset of the other can be that the number of nodes in the first feature point topology is less than the number of nodes in the second feature point topology, and after removing the extra nodes in the second feature point topology, it has the same structure as the first feature point topology, and vice versa.

[0073] Further, the steps of visually presenting the complete traceability information of the agricultural product in combination with the photo of the agricultural product specifically include: Obtain the node traceability information of the agricultural product at each traceability node in the complete traceability information of the agricultural product, where the traceability node information includes the photo of the agricultural product reported by the corresponding traceability node and its visual gradient feature; Present the photos of the agricultural product at each traceability node in chronological order; Fuse and present the visual gradient features of the agricultural product at each traceability node on the photo corresponding to the traceability node.

[0074] Specifically, the traceability node refers to the merchant node that reports the traceability information of the agricultural product to the agricultural product traceability server each time, and it can be a producer, processor, inspection contractor, carrier, or distributor, etc. The node traceability information at least includes information such as the scanning time, scanning location, photo of the agricultural product, visual gradient feature, and merchant ID.

[0075] Further, the steps of fusing and presenting the visual gradient features of the agricultural product at each traceability node on the photo corresponding to the traceability node specifically include: Determine the photo of the agricultural product and the visual gradient feature in the node traceability information of the same traceability node as the target photo and the target gradient feature respectively; Traverse each feature point in the target gradient feature to obtain the gradient feature value and gradient feature direction of the corresponding feature point; Obtain the pixel coordinates of each feature point in the target gradient feature on the target photo; Draw a circle on the target photo with the pixel coordinates as the center and the gradient feature value as the radius; Draw a feature line corresponding to the feature point on the target photo with the pixel coordinates as the starting point and the central angle of the gradient feature direction as the extension direction, and the end point of the feature line falls on the circle.

[0076] It should be noted that, in this text, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprising", "including" or any other variant thereof are intended to cover non-exclusive inclusion, such that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or device comprising the said element.

[0077] As described above with reference to the embodiments of the present invention, these embodiments do not describe all the details in detail, nor do they limit the invention to the specific embodiments described. Obviously, many modifications and variations can be made based on the above description. The purpose of selecting and specifically describing these embodiments in this specification is to better explain the principles and practical applications of the present invention, so that those skilled in the art can make good use of the present invention and its modifications. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. A visual tracking system for agricultural products, characterized in that: The invention comprises a camera device for taking photos of agricultural products, a barcode recognition device for capturing and recognizing a two-dimensional barcode on an outer package of agricultural products, an agricultural product tracking server for storing complete tracking information of agricultural products, and a visual tracking client connected to the camera device, the barcode recognition device and the agricultural product tracking server, wherein the visual tracking client is configured as follows: Obtaining a photo of the agricultural product and a tracking information label, wherein the tracking information label is a text or graphic label printed on the outer packaging of the agricultural product and stores the tracking information of the agricultural product; Parsing the tracking information tag to extract basic tracking information of the agricultural product from the tracking information tag, the basic tracking information including a unique identification of the agricultural product and a first visual gradient feature; determining whether a first visual gradient feature in the tracking information label matches the photograph of the agricultural product; When the first visual gradient feature in the tracking information tag matches the photo of the agricultural product, obtaining complete tracking information of the agricultural product from an agricultural product tracking server based on the unique identification of the agricultural product; The complete tracking information of the agricultural products is visually presented in combination with the photos of the agricultural products.

2. A visual tracking method for agricultural products, characterized in that: include: Obtaining a photo of the agricultural product and a tracking information label, wherein the tracking information label is a text or graphic label printed on the outer packaging of the agricultural product and stores the tracking information of the agricultural product; Parsing the tracking information tag to extract basic tracking information of the agricultural product from the tracking information tag, the basic tracking information including a unique identification of the agricultural product and a first visual gradient feature; determining whether a first visual gradient feature in the tracking information label matches the photograph of the agricultural product; When the first visual gradient feature in the tracking information tag matches the photo of the agricultural product, obtaining complete tracking information of the agricultural product from an agricultural product tracking server based on the unique identification of the agricultural product; The complete tracking information of the agricultural products is visually presented in combination with the photos of the agricultural products.

3. The visual tracking method of agricultural products according to claim 2, characterized in that: Before you get into the steps of getting photos of produce and labels with tracking information, include: before packaging the agricultural products, taking photos of the agricultural products; extracting a first visual gradient feature of the agricultural product from the photo of the agricultural product; generating a unique identification mark for the agricultural product; Writing the unique identification mark and the first visual gradient feature of the agricultural product into the tracking information label of the agricultural product; The tracking information label is printed or attached to the outer packaging of the agricultural product.

4. The visual tracking method of agricultural products according to claim 3, characterized in that: The step of extracting the first visual gradient feature of the agricultural product from the photo of the agricultural product specifically includes: configuring recognition scales for extracting key recognition features; extracting a main image of the agricultural product from the photo of the agricultural product; Performing denoising and smoothing processing on the subject image; Finding local extreme value points in the subject image subjected to denoising and smoothing based on the identification scale; Determining the local extreme point as a feature point of the subject image; A first visual gradient feature of the agricultural product at the feature point is extracted from the main image.

5. The visual tracking method of agricultural products according to claim 4, characterized in that: The step of extracting the first visual gradient feature of the agricultural product at the local extreme point specifically includes: configuring a gradient statistical scale of a first visualized gradient feature for statistically analyzing feature points, wherein the first visualized gradient feature includes a gradient feature value and a gradient feature direction of the feature point; Taking each feature point on the subject image as the center of the circle and the gradient statistical scale as the radius, determine the gradient statistical scale range of the corresponding feature point; Calculating the gradient feature value and gradient feature direction of the feature point according to the gradient features of each pixel point within the gradient statistical scale range of the feature point; The gradient feature value and gradient feature direction at each feature point on the main image are determined as the first visual gradient feature of the agricultural product.

6. The visual tracking method of agricultural products according to claim 5, characterized in that: The step of calculating the gradient feature value and gradient feature direction of the feature point according to the gradient features of each pixel point within the gradient statistical scale range of the feature point specifically includes: Traversing each feature point on the subject image; Determine the traversed feature points as target feature points; Traverse the gradient statistical scale range of the target feature point and the gradient amplitude of each pixel and the gradient direction ,in and are the horizontal coordinate and the vertical coordinate of the corresponding pixel on the subject image respectively; Count the gradient direction of the target feature point within the gradient statistics scale and fall into each angle interval The number of pixels ,in 1 to A positive integer between 360 degrees divided into preset is the number of angle intervals of the interval length, and satisfy: ； Calculate each angle interval The cumulative gradient amplitude of : ； in 1 to A positive integer between is the first The horizontal coordinate of the pixel, is the first The vertical coordinate of the pixel; In 1 to Determine a positive integer between , so that the gradient direction falls into the angle interval The cumulative gradient amplitude satisfy: ； The first The cumulative gradient amplitude of the angle interval Determine the gradient feature value of the target feature point; The first Angle range The gradient feature direction of the target feature point is determined.

7. The visual tracking method of agricultural products according to claim 2, characterized in that: The step of determining whether the first visual gradient feature in the tracking information label matches the photo of the agricultural product specifically includes: extracting a second visual gradient feature of the agricultural product from the photo of the agricultural product; comparing the first visual gradient feature with the second visual gradient feature to determine whether the second visual gradient feature is the same as the first visual gradient feature; When the second visual gradient feature is the same as the first visual gradient feature, it is determined that the first visual gradient feature in the tracking information label matches the photo of the agricultural product.

8. The visual tracking method of agricultural products according to claim 7, characterized in that: The step of comparing the first visualized gradient feature with the second visualized gradient feature to determine whether the second visualized gradient feature is the same as the first visualized gradient feature specifically includes: Get the number of feature points of the first visual gradient feature And the number of feature points of the second visualized gradient feature ; Calculate the number of feature points of the first visual gradient feature And the number of feature points of the second visualized gradient feature The difference in the number of feature points: ； When the number of feature points is different When the value is less than or equal to a preset deviation threshold, a feature point correspondence relationship between the first visualized gradient feature and the second visualized gradient feature is established; Comparing the gradient feature directions between feature points having a corresponding relationship in the first visualized gradient feature and the second visualized gradient feature; When the gradient directions of the feature points having a corresponding relationship in the first visualized gradient feature and the second visualized gradient feature are in the same angle interval, calculating the gradient feature value deviation between the feature points having a corresponding relationship in the first visualized gradient feature and the second visualized gradient feature; When a gradient feature value deviation between feature points having a corresponding relationship in the first visualized gradient feature and the second visualized gradient feature is less than or equal to a preset deviation threshold, it is determined that the second visualized gradient feature is the same as the first visualized gradient feature.

9. The visual tracking method of agricultural products according to claim 8, characterized in that: The first visualized gradient feature and the second visualized gradient feature include pixel coordinates of each feature point, and the step of establishing a corresponding relationship between feature points of the first visualized gradient feature and the second visualized gradient feature specifically includes: Constructing a first feature point topology based on the pixel coordinates of the feature points included in the first visualized gradient feature and constructing a second feature point topology based on the pixel coordinates of the feature points included in the second visualized gradient feature; comparing the topological shapes of the first feature point topology and the second feature point topology; Determining whether the topological shapes of the first feature point topology and the second feature point topology are the same; When the topological shapes of the first feature point topology and the second feature point topology are different, determining whether the topological shapes of the first feature point topology and the second feature point topology satisfy that one is a subset of the other; When the topological shapes of the first feature point topology and the second feature point topology are the same, or the topological shapes of the first feature point topology and the second feature point topology satisfy that one is a subset of the other, determining that feature points of the first visualization gradient feature and the second visualization gradient feature have a corresponding relationship; A correspondence relationship between feature points at corresponding positions in the first feature point topology and the second feature point topology is established.

10. The visual tracking method of agricultural products according to claim 2, characterized in that: The step of visually presenting the complete tracking information of the agricultural product in combination with the photo of the agricultural product specifically includes: Obtaining node tracking information of the agricultural product in each tracking node from the complete tracking information of the agricultural product, wherein the tracking node information includes a photo of the agricultural product reported by the corresponding tracking node and its visual gradient feature; Presenting photos of the agricultural products at each tracking point in chronological order; The visualized gradient features of the agricultural products at each tracking node are fused and presented on a photo of the corresponding tracking node.

Images