Image Processing Method, Device, System and Storage Medium Based on Mesh Spot

Abstract

The present invention discloses an image processing method, device, system and storage medium based on a reticular object spot. The method includes the following steps: obtaining a discrete spot image formed by a line laser beam hitting a target reticular object; wherein, the discrete spot image includes at least four segments of laser spots; identifying the coverage area of each laser point in the discrete spot image and extracting the centroid coordinates of each laser point based on this; selecting at least four target areas from the discrete spot image; wherein, the target areas correspond to the laser spots one by one, and each target area contains a plurality of laser points; performing linear fitting on the centroid coordinates in each target area to obtain a fitting line for judging the deflection situation and distance of the target reticular object. The image processing method based on the reticular object spot disclosed by the present invention can solve the technical problem that it is difficult to form an effective input condition for distance measurement by collecting discrete spots on the reticular object at present, thereby causing an obstacle to determining the actual distance of the reticular object.

Description

Technical Field

[0001] The present invention belongs to the technical field of image processing, and particularly relates to an image processing method based on a reticular object light spot, an image processing device based on a reticular object light spot, an image processing system based on a reticular object light spot, and a computer-readable storage medium. Background Art

[0002] Currently, the fishery aquaculture in China has a development trend from inshore to deep sea. During the maintenance operation of the aquaculture net, it is necessary to obtain the distance between the operation mechanism and the net in real time to avoid entanglement between the operation mechanism and the net and ensure the safe and smooth progress of the net maintenance operation.

[0003] Regarding the ranging between the operation mechanism and the net, it is mainly realized by an underwater laser rangefinder. The traditional underwater laser rangefinder is a single-point laser rangefinder based on the phase-type laser ranging principle. However, reticular objects such as nets have a periodic thin wire and mesh structure, and have the characteristic of a small proportion of the acousto-optic signal reflection area, that is, the target plane of the net is periodically hollow, and the point laser beam emitted by the underwater laser rangefinder may pass through the hollow area of the mesh and fail to obtain a reflection signal.

[0004] Based on this, a line laser rangefinder needs to be used to measure the distance of the reticular object. However, there are complex ocean currents such as advection, turbulence, and internal waves in the aquaculture water body in the deep-sea environment, and the flexible structure of the reticular object makes it have transient characteristics that fluctuate with the ocean current; due to the complex fluctuation conditions of the reticular object, the discrete light spots formed by the line laser beam hitting the reticular object are also difficult to find rules, and it is difficult to form effective input conditions for subsequent mathematical solution processes through the collected discrete light spots, which poses a great obstacle to finally determining the actual distance of the reticular object. Summary of the Invention

[0005] In order to overcome the above-mentioned shortcomings of the prior art, the purpose of the present invention is to provide an image processing method based on a reticular object light spot, aiming to solve the technical problem that it is currently difficult to form effective input conditions for ranging by collecting discrete light spots on the reticular object, thereby posing an obstacle to determining the actual distance of the reticular object.

[0006] The technical solution adopted by the present invention to achieve its purpose is as follows:

[0007] An image processing method based on a reticular object light spot, the image processing method based on a reticular object light spot includes the following steps:

[0008] Obtain a discrete light spot image formed by a line laser beam hitting a target reticular object; wherein, the discrete light spot image includes at least four laser light spots;

[0009] Identify the coverage area of each laser point in the discrete light spot image;

[0010] Extract the centroid coordinates of each laser point based on the coverage area of each laser point;

[0011] Select at least four target regions from the discrete spot image; wherein, each target region corresponds to a laser spot one by one, and each target region contains a plurality of the laser points;

[0012] Perform linear fitting on the centroid coordinates in each target region to obtain a fitting line for judging the deflection situation and distance of the target mesh.

[0013] Further, before the step of identifying the coverage area of each laser point in the discrete spot image, the method further includes:

[0014] Perform denoising processing on the discrete spot image.

[0015] Further, before the step of identifying the coverage area of each laser point in the discrete spot image, the method further includes:

[0016] Perform grayscale processing on the discrete spot image;

[0017] Based on a grayscale threshold, perform binary processing on the grayscale processed discrete spot image.

[0018] Further, the step of selecting at least four target regions from the discrete spot image specifically includes:

[0019] Set at least four classifiers based on the linear relationship of the laser spots; wherein, each classifier corresponds to a laser spot one by one;

[0020] The step of performing linear fitting on the centroid coordinates in each target region to obtain a fitting line for judging the deflection situation and distance of the target mesh specifically includes:

[0021] Adopt a density clustering algorithm to classify and import the centroid coordinates into the corresponding classifiers;

[0022] Perform linear fitting on the centroid coordinates stored in each classifier to obtain the fitting line for judging the deflection situation and distance of the target mesh.

[0023] Further, the step of identifying the coverage area of each laser point in the discrete spot image specifically includes:

[0024] Adopt a Canny edge detection algorithm to identify the coverage area of each laser point in the discrete spot image.

[0025] Further, the step of denoising the discrete spot image specifically includes:

[0026] Performing denoising processing on the discrete spot image by using a Gaussian filtering method.

[0027] Further, the step of binarizing the gray-scaled discrete spot image based on a gray threshold specifically includes:

[0028] Performing binarization processing on the gray-scaled discrete spot image by using the maximum inter-class variance method.

[0029] Correspondingly, the present invention further provides an image processing device based on a reticular object spot. The image processing device based on a reticular object spot includes:

[0030] An acquisition module, configured to acquire a discrete spot image formed by a line laser beam hitting a target reticular object;

[0031] An identification module, configured to identify the coverage area of each laser point in the discrete spot image;

[0032] An extraction module, configured to extract the centroid coordinates of each laser point according to the coverage area;

[0033] A classification module, configured to select at least four target areas from the discrete spot image;

[0034] A fitting module, configured to perform linear fitting on the centroid coordinates in each target area to obtain a fitting line for judging the deflection situation and distance of the target reticular object.

[0035] Correspondingly, the present invention further provides an image processing system based on a reticular object spot. The image processing system based on a reticular object spot includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the computer program is executed by the processor, the steps of the image processing method based on a reticular object spot as described above are implemented.

[0036] Correspondingly, the present invention further provides a computer-readable storage medium. An image processing program based on a reticular object spot is stored on the computer-readable storage medium. When the image processing program based on a reticular object spot is executed by a processor, the steps of the image processing method based on a reticular object spot as described above are implemented.

[0037] Compared with the prior art, the beneficial effects of the present invention are:

[0038] The image processing method based on the mesh spot proposed by the present invention emits multiple laser beams onto the target mesh to form multiple segments of laser spots, then acquires the discrete spot image containing these multiple segments of laser spots, selects the corresponding target regions on the discrete spot image according to the number of laser spots, each target region contains all the laser points in the corresponding laser spot, and extracts the centroid coordinates of the laser points. Finally, linear fitting is performed on the centroid coordinates of the laser points contained in each target region to obtain multiple fitting lines, so that the deflection of the target mesh can be judged through these multiple fitting lines and the actual distance from the vision sensing device to the target mesh can be deduced. Thus, the acquired discrete spot image is converted into an effective input condition for ranging, realizing accurate underwater laser ranging for flexible meshes. Description of the Drawings

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained according to the structures shown in these drawings.

[0040] Figure 1 It is a schematic flowchart of an embodiment of the image processing method based on the mesh spot of the present invention;

[0041] Figure 2 It is a schematic detailed flowchart of an embodiment of the image processing method based on the mesh spot of the present invention;

[0042] Figure 3 It is a schematic structural diagram of the ranging device used in the embodiment solution of the present invention;

[0043] Figure 4 It is a schematic diagram of the discrete spot image in the embodiment solution of the present invention;

[0044] Figure 5 It is a schematic diagram of the fitting lines generated from the discrete spot image in the embodiment solution of the present invention;

[0045] Figure 6 It is a schematic structural diagram of the device involved in the embodiment solution of the present invention;

[0046] Figure 7 It is a schematic system structure diagram of the hardware operating environment involved in the embodiment solution of the present invention.

[0047] Description of the Reference Numerals:

[0048] Label Name Label Name 1 Laser emission device 2 Visual sensing device

[0049] The implementation, functional features and advantages of the present invention will be further described in conjunction with the embodiments with reference to the accompanying drawings. Detailed implementation manners

[0050] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0051] As Figure 7 shown, Figure 7 is a schematic structural diagram of an image processing system based on a reticular object spot involved in the embodiment solution of the present invention.

[0052] As Figure 7 shown, the image processing system based on the reticular object spot may include: a processor 1001, such as a CPU, a network interface 1004, a user interface 1003, a memory 1005, and a communication bus 1002. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display) and an input unit such as a keyboard (Keyboard). Optionally, the user interface 1003 may further include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface). The memory 1005 may be a high-speed RAM memory or a stable memory (non-volatile memory), such as a disk memory. Optionally, the memory 1005 may also be a storage device independent of the foregoing processor 1001.

[0053] Optionally, the image processing system based on the reticular object spot may further include a camera, an RF (Radio Frequency) circuit, sensors, an audio circuit, a WiFi module, and so on. Among them, the sensors may include a light sensor, a motion sensor, an infrared sensor, and other sensors, which will not be elaborated here.

[0054] Those skilled in the art can understand that Figure 7 the specific structure shown in

[0055] As Figure 7 shown, the memory 1005, as a computer-readable storage medium, may include an operating system, a network communication module, a user interface module, and an image processing program based on the reticular object spot.

[0056] In Figure 7 the shown image processing system based on reticular spot, the network interface 1004 is mainly used to connect to the background server and conduct data communication with the background server; the user interface 1003 is mainly used to connect to the client (user side) and conduct data communication with the client; and the processor 1001 can be used to call the image processing program based on reticular spot stored in the memory 1005 and perform the following operations:

[0057] Obtain the discrete spot image formed by the line laser beam hitting the target reticular object; wherein, the discrete spot image includes at least four segments of laser spots;

[0058] Identify the coverage area of each laser point in the discrete spot image;

[0059] Based on the coverage area of each laser point, extract the centroid coordinates of each laser point;

[0060] Select at least four target areas from the discrete spot image; wherein, the target areas correspond to the laser spots one by one, and each target area contains multiple laser points;

[0061] Perform linear fitting on the centroid coordinates in each target area to obtain a fitting line for judging the deflection situation and distance of the target reticular object.

[0062] Furthermore, the processor 1001 can call the network operation control application program stored in the memory 1005 and also perform the following operations:

[0063] Perform denoising processing on the discrete spot image.

[0064] Furthermore, the processor 1001 can call the network operation control application program stored in the memory 1005 and also perform the following operations:

[0065] Perform grayscale processing on the discrete spot image;

[0066] Based on the grayscale threshold, perform binary processing on the grayscale processed discrete spot image.

[0067] Furthermore, the processor 1001 can call the network operation control application program stored in the memory 1005 and also perform the following operations:

[0068] Set at least four classifiers based on the linear relationship of the laser spots; wherein, each classifier corresponds to each laser spot one by one.

[0069] Furthermore, the processor 1001 can call the network operation control application program stored in the memory 1005 and also perform the following operations:

[0070] Use the density clustering algorithm to classify and import the centroid coordinates into the corresponding classifier;

[0071] Perform linear fitting on the centroid coordinates stored in each classifier to obtain a fitting line for judging the deflection situation and distance of the target mesh.

[0072] Further, the processor 1001 can call the network operation control application program stored in the memory 1005 and also perform the following operations:

[0073] Use the Canny edge detection algorithm to identify the coverage area of each laser point in the discrete spot image.

[0074] Further, the processor 1001 can call the network operation control application program stored in the memory 1005 and also perform the following operations:

[0075] Use the Gaussian filtering method to denoise the discrete spot image.

[0076] Further, the processor 1001 can call the network operation control application program stored in the memory 1005 and also perform the following operations:

[0077] Use the maximum inter-class variance method to binarize the discrete spot image after grayscale processing.

[0078] Refer to Figure 1 , an embodiment of the present invention provides an image processing method based on a mesh spot. The image processing method based on a mesh spot includes:

[0079] S1, obtain a discrete spot image formed by a line laser beam hitting a target mesh; wherein, the discrete spot image includes at least four segments of laser spots;

[0080] The target mesh can specifically be a flexible member woven from wire meshes and having multiple mesh holes, such as a net for fishery culture; the line laser beam can be regarded as a continuous line segment composed of several laser points. No matter how the target mesh fluctuates in the water body, there are always laser points on the continuous line laser beam that can hit the wire meshes of the target mesh and be collected by the visual sensing device. Among them, the visual sensing device can be an underwater camera communicatively connected to the image processing device. After the optical main plane of the underwater camera captures several laser points hitting the wire meshes of the target mesh, they are acquired by the acquisition module of the image processing device and displayed on the imaging plane in the form of a discrete spot image. The discrete spots in the discrete spot image are the set of laser points hitting the wire meshes of the target mesh on the image.

[0081] Figure 3The structure of a ranging device is shown. As shown in the figure, four laser emission devices are arranged around the side wall of the visual sensing device in a square layout, and the four laser emission devices are respectively located at the midpoints of the sides of the square. In this way, a number of light spots formed by the four line laser beams emitted by the four laser emission devices hitting the target mesh wires are acquired by the acquisition module, and then four segments of discrete light spots can be formed in the discrete light spot image, and each segment of discrete light spot corresponds one-to-one to each line laser beam. Similarly, when the number of laser emission devices is set to more than four, the corresponding number of discrete light spots can also be formed in the discrete light spot image.

[0082] Specifically, referring to Figure 2 , before step S2, it includes:

[0083] S201, perform denoising processing on the discrete light spot image;

[0084] Specifically, step S201 specifically includes:

[0085] S2011, perform denoising processing on the discrete light spot image using the Gaussian filtering method;

[0086] Due to reasons such as the existence of backscattering of laser in water and high water turbidity, the discrete light spot image collected by the visual sensing device and displayed on the image processing device has problems such as low contrast and blurred details caused by low signal-to-noise ratio. Therefore, it is first necessary to perform denoising preprocessing on the discrete light spot image. Specifically, the methods of denoising preprocessing include, but are not limited to, median filtering, mean filtering, and Gaussian filtering, etc. Here, it is preferably to use the Gaussian filtering denoising method. Among them, the size of the Gaussian filter kernel can be set to 3×3. Gaussian filtering is a local linear smoothing filter, which is suitable for eliminating Gaussian noise. The specific process of using Gaussian filtering to eliminate image noise is: scan each pixel in the image with a template, and then replace the value of the template center pixel with the weighted average value of the pixels in the area determined by the template. After performing denoising processing on the discrete light spot image, the effective pixels of the discrete light spot can be effectively expanded.

[0087] Specifically, referring to Figure 2 , before step S2, it also includes:

[0088] S202, perform grayscale processing on the discrete light spot image;

[0089] S203, perform binarization processing on the grayscale processed discrete light spot image based on the grayscale threshold.

[0090] After the discrete spot image is denoised, it can be grayed. Graying is to convert the initially colored discrete spot image into a gray image. Specifically, the image processing device usually uses the RGB model (additive color mixing model) for image display. The RGB model uses red (R), green (G), and blue (B) as the three primary colors of the image. The value range of R, G, and B is 0 to 255. When R=G=B, the image presents a gray color. Graying is to adjust the R, G, and B values ​​of each pixel in the discrete spot image to a common gray value according to a certain ratio. The gray value can be calculated by the following formula:

[0091]

[0092] The discrete spot image after grayscale processing is binarized, and a grayscale threshold can be pre-set. The grayscale threshold value ranges from 0 to 255, and it can be set that when the grayscale value of a pixel in the discrete spot image is higher than the grayscale threshold, the pixel is converted to black; when the grayscale value of a pixel in the discrete spot image is lower than the grayscale threshold, the pixel is converted to white. In this way, the discrete spot image after binarization processing becomes a black and white image, which can highlight the position of the laser point in the image and avoid affecting the subsequent discrimination of the laser point due to the mixed color of the image.

[0093] Specifically, step S203 includes:

[0094] S2031, using the maximum inter-class variance method to perform binarization processing on the discrete light spot image after grayscale processing.

[0095] In practical applications, when performing binarization, methods for determining grayscale thresholds include but are not limited to the maximum inter-class variance method (OSTU algorithm), grayscale mean method, and adaptive threshold method. The maximum inter-class variance method (OSTU algorithm) is preferably used here, which can make the binarization process for discrete spot images more efficient. Among them, the maximum inter-class variance method (OSTU algorithm) is a commonly used algorithm in this field, and its calculation process is not specifically described here.

[0096] S2, identifying the coverage area of ​​each laser point in the discrete light spot image;

[0097] Specifically, refer to Figure 2 , step S2 specifically includes:

[0098] S21, using the Canny edge detection algorithm to identify the coverage area of ​​each laser point in the discrete spot image;

[0099] Identify the coverage area of the laser points, that is, determine the closed contour of the laser points. Specifically, methods such as the Sobel edge detection operator, Canny edge detection algorithm, gradient edge detection, Roberts edge detection operator, Laplace edge detection operator, LoG operator, or differential edge detection can be used, all of which can achieve the technical effect of determining the closed contour of the laser points. Here, the Canny edge detection algorithm is preferably used.

[0100] Through image edge detection, the amount of data is greatly reduced, and irrelevant information can be removed. Especially for laser points with scattering in water, their important structural attributes are retained for subsequent further image processing based on the laser points.

[0101] S3. Based on the coverage area of each laser point, extract the centroid coordinates of each laser point.

[0102] The centroid of a laser point can be considered as the point that can represent the actual geometric center position of the laser point. In the specific implementation process, for the closed contour of the laser points generated by the edge detection algorithm, the centroid coordinates of the laser points can be calculated through relevant image contour centroid algorithms, such as by using the cvFindContours function.

[0103] S4. Select at least four target regions from the discrete spot image; where the target regions correspond one-to-one with the laser spots, and each target region contains multiple laser points.

[0104] If a ranging device as shown in Figure 3 is used, four laser spots as shown in Figure 4 will be generated in the discrete spot image, and these four laser spots enclose a quadrilateral. At this time, the number of target regions should be four corresponding to the laser spots, specifically including two vertical quadrilateral (or strip) regions and two horizontal quadrilateral (or strip) regions, as shown in Figure 4 . The two vertical regions correspond to two vertical laser spots respectively and should contain all the laser points in these two vertical laser spots. The two horizontal regions correspond to two horizontal laser spots respectively and should contain all the laser points in these two horizontal laser spots. In the specific implementation process, the selection of the target regions can be achieved by using the ROI (region of interest) extraction algorithm. After selecting the target regions through the ROI extraction algorithm, other regions in the discrete spot image can be automatically ignored, thereby reducing the computing intensity of the system, saving power consumption, and improving the system computing efficiency in the subsequent solution process.

[0105] Specifically, referring to Figure 2 , step S4 specifically includes:

[0106] S41. Set at least four classifiers based on the linear relationship of the laser spots; where each classifier corresponds to each laser spot one by one.

[0107] The linear relationship of the laser spots refers to the linear distribution law of four segments of laser spots in the discrete spot image. Each classifier corresponds to a linear distribution law (the linear distribution law can be preset in the classifier in the form of specific thresholds). Thus, the classifier can frame the corresponding target area from the discrete spot image based on this threshold. For any horizontal target area shown, it contains all the laser points in this segment of the horizontal laser spot. At the same time, the number of laser points in other segments of laser spots it contains is controlled within a certain range, which can avoid affecting the subsequent solution process for this segment of laser spots. The same applies to other laser spots. Figure 4 As described above, each classifier selects a target area (taking the example shown, the first classifier selects the first vertical area, the second classifier selects the second vertical area, the third classifier selects the first horizontal area, and the fourth classifier selects the second horizontal area).

[0108] As mentioned above, each classifier selects a target area (taking the example shown, the first classifier selects the first vertical area, the second classifier selects the second vertical area, the third classifier selects the first horizontal area, and the fourth classifier selects the second horizontal area). Figure 4 As shown in the figure, the first classifier selects the first vertical area, the second classifier selects the second vertical area, the third classifier selects the first horizontal area, and the fourth classifier selects the second horizontal area).

[0109] S5. Perform linear fitting on the centroid coordinates in each of the target areas to obtain a fitting line for judging the deflection situation and distance of the target mesh.

[0110] Linear fitting can convert discrete laser points into continuous line segments (i.e., fitting lines) that can reflect their laws and functional relationships. The fitting lines can characterize the state of the corresponding line laser beam hitting the target mesh. Specifically, the relative positions between multiple fitting lines will change due to the different distances from the points where the line laser beam hits the target mesh to the vision sensing device. When the target mesh has different deflection situations and different distances from the vision sensing device, the multiple fitting lines will be spaced at different distances according to the geometric relationship of laser projection and a certain imaging magnification relationship. Therefore, the deflection situation of the target mesh can be judged by the relative distance between the fitting lines, and the actual distance from the vision sensing device to the target mesh can be deduced based on the geometric relationship of laser projection, thereby realizing accurate underwater ranging for the flexible mesh.

[0111] Specifically, referring to Figure 2 , step S5 specifically includes:

[0112] S51. Use the density clustering algorithm to classify and import the centroid coordinates into the corresponding classifiers;

[0113] S52. Perform linear fitting on the centroid coordinates stored in each classifier to obtain a fitting line for judging the deflection situation and distance of the target mesh.

[0114] TakingFigure 4 Taking the case where there are four target areas shown as an example, based on the four set classifiers, the density clustering algorithm (DBSCAN) can divide the laser spots into four categories according to the linear distribution law, and import the centroid coordinates into the corresponding classifiers (that is, import the centroid coordinates of the laser points in the target area corresponding to each classifier into this classifier). Then, perform linear fitting on the centroid coordinates of the laser points in each classifier. Finally, the four fitting lines as shown in Figure 5 will be obtained. By measuring the relative distances between these four fitting lines, the deflection situation of the target mesh can be judged and the actual distance from the visual sensing device to the target mesh can be deduced. Among them, linear fitting is a commonly used technology in this field, and the solution process thereof will not be specifically described here.

[0115] It can be seen that the image processing method based on the mesh spot provided in this embodiment emits multiple line laser beams onto the target mesh to form multiple segments of laser spots, then obtains the discrete spot image containing these multiple segments of laser spots, selects the corresponding target areas on the discrete spot image according to the number of laser spots, each target area contains all the laser points in the laser spot corresponding to it, and extracts the centroid coordinates of the laser points. Finally, perform linear fitting on the centroid coordinates of the laser points contained in each target area to obtain multiple fitting lines, so that the deflection situation of the target mesh can be judged through these multiple fitting lines and the actual distance from the visual sensing device to the target mesh can be deduced. Thus, the obtained discrete spot image is converted into an effective input condition for ranging, realizing accurate underwater laser ranging for the flexible mesh.

[0116] Correspondingly, referring to Figure 6 , the embodiment of the present invention also provides an image processing device based on the mesh spot. The image processing device based on the mesh spot includes:

[0117] An acquisition module 10 for acquiring the discrete spot image formed by the line laser beam hitting the target mesh;

[0118] An identification module 20 for identifying the coverage area of each laser point in the discrete spot image;

[0119] An extraction module 30 for extracting the centroid coordinates of each laser point according to the coverage area;

[0120] A classification module 40 for selecting at least four target areas from the discrete spot image;

[0121] A fitting module 50 for performing linear fitting on the centroid coordinates in each target area to obtain the fitting lines for judging the deflection situation and distance of the target mesh.

[0122] The image processing device based on the reticular spot in this embodiment is used to implement the aforementioned image processing method based on the reticular spot. Therefore, the specific implementation manners in the image processing device based on the reticular spot can be seen in the embodiment part of the image processing method based on the reticular spot in the foregoing text. For example, the acquisition module 10, the recognition module 20, the extraction module 30, the classification module 40, and the fitting module 50 are respectively used to implement steps S1, S2, S3, S4, and S5 in the aforementioned image processing method based on the reticular spot. Therefore, the specific implementation manners can refer to the description of the above embodiments and will not be elaborated herein.

[0123] Correspondingly, an embodiment of the present invention further provides a computer-readable storage medium, on which an image processing program based on the reticular spot is stored. When the image processing program based on the reticular spot is executed by a processor, the steps of the image processing method based on the reticular spot in any of the above embodiments are implemented.

[0124] In this embodiment, the above computer-readable storage medium may include, but is not limited to, any type of disk (including floppy disk, hard disk, optical disk, CD-ROM, and magneto-optical disk), ROM (Read-Only Memory), RAM (Random Access Memory), EPROM (Erasable Programmable Read-Only Memory), EEPROM (Electrically Erasable Programmable Read-Only Memory), flash memory, magnetic card, or optical card, etc., various media that can store program codes.

[0125] It should be noted that in this article, the term "including", "comprising", or any other variant thereof is intended to cover a non-exclusive inclusion, so that a process, method, article, or system including a series of elements not only includes those elements, but also includes other elements that are not explicitly listed, or further includes elements inherent to such process, method, article, or system. Without more limitations, an element defined by the statement "including a..." does not exclude the existence of additional identical elements in the process, method, article, or system including that element.

[0126] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages and disadvantages of the embodiments.

[0127] Through the description of the above embodiments, those skilled in the art can clearly understand that the above-described example methods can be implemented by means of software plus a necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art can be embodied in the form of a software product. This computer software product is stored in a storage medium as described above (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions for causing a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in various embodiments of the present invention.

[0128] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the specification and drawings of the present invention, or directly or indirectly applied in other related technical fields, shall be equally included in the patent protection scope of the present invention.

Claims

1. An image processing method based on a reticular light spot, characterized in that, the image processing method based on the reticular light spot includes the following steps: Obtain a discrete light spot image formed by a line laser beam hitting a target reticular object; wherein, the discrete light spot image includes at least four segments of laser light spots; Identify the coverage area of each laser point in the discrete light spot image; Based on the coverage area of each laser point, extract the centroid coordinates of each laser point; Select at least four target areas from the discrete light spot image; wherein, the target areas correspond to the laser light spots one by one, and each target area contains a plurality of the laser points; Perform linear fitting on the centroid coordinates in each target area to obtain a fitting line for judging the deflection situation and distance of the target reticular object; the distance is the distance between the visual sensing device and the target reticular object.

2. The image processing method based on the reticular light spot according to claim 1, characterized in that, before the step of identifying the coverage area of each laser point in the discrete light spot image, it further includes: Perform denoising processing on the discrete light spot image.

3. The image processing method based on the reticular light spot according to claim 1, characterized in that, before the step of identifying the coverage area of each laser point in the discrete light spot image, it further includes: Perform grayscale processing on the discrete light spot image; Based on a grayscale threshold, perform binary processing on the grayscale-processed discrete light spot image.

4. The image processing method based on the reticular light spot according to claim 1, characterized in that, the step of selecting at least four target areas from the discrete light spot image specifically includes: Set at least four classifiers based on the linear relationship of the laser light spots; wherein, each classifier corresponds to each laser light spot one by one; the step of performing linear fitting on the centroid coordinates in each target area to obtain a fitting line for judging the deflection situation and distance of the target reticular object specifically includes: Use the density clustering algorithm to classify and import the centroid coordinates into the corresponding classifiers; Perform linear fitting on the centroid coordinates stored in each classifier to obtain the fitting line for judging the deflection situation and distance of the target reticular object.

5. The image processing method based on the reticular light spot according to claim 1, characterized in that, the step of identifying the coverage area of each laser point in the discrete light spot image specifically includes: Use the Canny edge detection algorithm to identify the coverage area of each laser point in the discrete light spot image.

6. The image processing method based on the reticular light spot according to claim 2, characterized in that, the step of performing denoising processing on the discrete light spot image specifically includes: Use the Gaussian filtering method to perform denoising processing on the discrete light spot image.

7. The image processing method based on the reticular light spot according to claim 3, characterized in that, the step of performing binary processing on the grayscale-processed discrete light spot image based on the grayscale threshold specifically includes: The maximum inter-class variance method is used to perform binarization processing on the grayscale discrete spot image.

8. An image processing device based on a reticular object spot, characterized in that the image processing device based on a reticular object spot includes: an acquisition module, configured to acquire a discrete spot image formed by a line laser beam hitting a target reticular object; an identification module, configured to identify the coverage area of each laser point in the discrete spot image; an extraction module, configured to extract the centroid coordinates of each laser point according to the coverage area; a classification module, configured to select at least four target areas from the discrete spot image; a fitting module, configured to perform linear fitting on the centroid coordinates in each target area to obtain a fitting line for judging the deflection condition and distance of the target reticular object; the distance is the distance between the vision sensing device and the target reticular object.

9. An image processing system based on a reticular object spot, characterized in that the image processing system based on a reticular object spot includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the computer program is executed by the processor, the steps of the image processing method based on a reticular object spot according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that an image processing program based on a reticular object spot is stored on the computer-readable storage medium. When the image processing program based on a reticular object spot is executed by a processor, the steps of the image processing method based on a reticular object spot according to any one of claims 1 to 7 are implemented.

Images