A steel coil positioning method, device, equipment and medium based on a cart

Abstract

The present disclosure provides a method, device, equipment, and medium for locating steel coils based on a cart. The method includes: collecting a non-filled light image of a cart, the cart being located in a parking space and loaded with horizontal steel coils, with a preset number of fill lights installed in different directions above the parking space; determining a target fill light and its light intensity from the preset number of fill lights based on the non-filled light image of the cart, and collecting fill light images of the cart based on the on / off status and light intensity of the target fill light, each fill light image including a bright area and a shadow area; fusing the fill light images to obtain a fused image; calculating the initial fitting coordinates and tilt angle of each steel coil based on the non-filled light image and the fused image, and calculating the center coordinates of each steel coil based on the initial fitting coordinates and the steel coil tilt angle. Rapid and accurate positioning of the steel coils is achieved by analyzing shadows after performing multi-directional fill light on the steel coils.

Description

Technical Field

[0001] The present disclosure relates to the technical field of steel coil scanning, and in particular to a steel coil positioning method, device, equipment and medium based on a pallet truck. Background Art

[0002] With the continuous advancement of industrial automation technology, the demand for efficient and safe operations in the steel coil transportation industry is increasing. Overhead crane systems (also known as automated material handling systems) have become an indispensable component of many modern steel coil warehouses. Compared to traditional manual operations, unmanned automated overhead crane systems not only significantly improve lifting efficiency and reduce labor costs, but also effectively reduce the risk of steel coils falling or being damaged, thereby significantly improving overall safety.

[0003] The steel coil positioning of the unmanned automatic overhead crane system generally adopts the laser scanning method. The laser scanning equipment obtains the point cloud data of the target area by emitting a laser beam to the target area and receiving the reflected signal, and calculates the specific position information of the steel coil based on the point cloud data. However, in actual application, due to factors such as ambient light interference, laser reflection angle deviation, and surface material characteristics of the steel coil, the quality of the point cloud data may be reduced, thereby affecting the positioning accuracy. In addition, the data processing of laser point cloud data is difficult and the calculation amount is high. The existence of these problems will affect the overall operating efficiency and reliability of the overhead crane system. Therefore, improving positioning accuracy and system reliability has become an urgent market demand. It is particularly important to propose a method that can quickly and accurately locate steel coils. Summary of the Invention

[0004] The present application provides a steel coil positioning method, device, equipment and medium based on a pallet truck to solve a series of problems in the prior art when positioning steel coils, such as complex data processing difficulty and large data volume, poor processing result accuracy, and slow positioning speed.

[0005] In a first aspect, the present disclosure provides a steel coil positioning method based on a pallet truck, the method comprising:

[0006] Capturing an image of a cart without fill light, the cart being located in a parking space and carrying a horizontally placed steel coil, wherein a preset number of fill lights are installed in different directions above the parking space;

[0007] Determining a target fill light and the light intensity of the target fill light from a preset number of fill lights based on the image of the vehicle without fill light, and collecting fill light images of the vehicle according to the on / off state and light intensity of the target fill light, wherein each fill light image includes a bright area and a shadow area;

[0008] fusing the fill-light images to obtain a fused image, wherein each steel coil on the truck exhibits at least a first shadow region and a second shadow region, wherein the first shadow region is located on the left side of the steel coil and the second shadow region is located on the right side of the steel coil;

[0009] The initial fitting coordinates and the tilt angle of each steel coil are calculated based on the non-filled light image and the fused image, and the center coordinates of each steel coil are calculated based on the initial fitting coordinates and the steel coil tilt angle.

[0010] According to a steel coil positioning method based on a pallet cart provided by the present disclosure, the initial fitting coordinates and the tilt angle of each steel coil are calculated based on the image without fill light and the fused image, and the center coordinates of each steel coil are calculated based on the initial fitting coordinates and the steel coil tilt angle, including: calculating the initial fitting coordinates of each steel coil on the pallet cart based on the fused image and the image without fill light; determining the feature information of the shadow area in each steel coil based on the fused image, and calculating the tilt angle of each steel coil based on the feature information of the shadow area in each steel coil; calculating the center coordinates of each steel coil based on the initial fitting coordinates and the steel coil tilt angle.

[0011] According to a steel coil positioning method based on a flatbed cart provided by the present disclosure, the characteristic information of the shadow area includes the position, shape, and size of the shadow area, and the calculation of the inclination angle of each steel coil based on the characteristic information of the shadow area in each steel coil at least includes: if the shadow area of ​​any steel coil is located on the left and right sides of the steel coil and the shadow area is a rectangle, then the inclination angle of the steel coil is calculated according to the width of the rectangle corresponding to the shadow area; if the shadow area of ​​any steel coil is located on the upper or lower side of the steel coil and the shadow area is an ellipse, then the inclination angle of the steel coil is calculated according to the minor semi-axis of the ellipse corresponding to the shadow area.

[0012] According to a steel coil positioning method based on a flatbed cart provided in the present disclosure, the inclination angle includes a roll angle and / or a pitch angle, and the center coordinates of each steel coil are calculated based on the initial fitting coordinates and the steel coil inclination angle, including: calculating a first offset of the steel coil center coordinates based on the roll angle of each steel coil, and / or calculating a second offset of the steel coil center coordinates based on the pitch angle of each steel coil; calculating the center coordinates of each steel coil based on the initial fitting coordinates and the first offset and / or the second offset.

[0013] According to a steel coil positioning method based on a pallet truck provided by the present disclosure, the fill-in light images are fused to obtain a fused image, including: constructing a mask for each fill-in light image, wherein the mask is 0 in the bright area of ​​the fill-in light image and 1 in the shadow area of ​​the fill-in light image; multiplying each fill-in light image by its own mask to obtain a shadow image corresponding to each fill-in light image; and superimposing the shadow image corresponding to each fill-in light image and any fill-in light image to obtain a fused image.

[0014] According to a steel coil positioning method based on a pallet truck provided by the present disclosure, the target fill light and the light intensity of the target fill light are determined from a preset number of fill lights based on the pallet truck's image without fill light, including: edge detection is performed on the pallet truck's image without fill light to determine the contour shape of each steel coil, if the contour shape of each steel coil is a rectangle, the target fill light includes at least fill lights in the left and right directions, otherwise the target fill light includes at least fill lights in the left, right, front and rear directions; the intensity of the ambient light is calculated based on the image without fill light; if the intensity of the ambient light is less than or equal to a first threshold, the light intensity of the target fill light is set to a preset light intensity value, otherwise, the target light intensity value of each target fill light is adaptively calculated based on the image without fill light.

[0015] According to a steel coil positioning method based on a pallet truck provided by the present disclosure, the target light intensity value of each target fill light is adaptively calculated based on the image without fill light, specifically including: for any target fill light, determining the initial light intensity of the target fill light based on the image without fill light; gradually adjusting the light intensity based on the initial light intensity and sequentially acquiring fill light images under various light intensities until the fill light image meets the fill light requirements, stopping adjusting the light intensity, and determining the light intensity that meets the fill light requirements as the target light intensity value of the target fill light.

[0016] In a second aspect, the present disclosure further provides a steel coil positioning device based on a plate cart, the device comprising:

[0017] a first acquisition module for acquiring an image of a cart without fill light, the cart being located in a parking space and carrying a horizontally placed steel coil, wherein a preset number of fill lights are installed in different directions above the parking space;

[0018] a second acquisition module, configured to determine a target fill light and the light intensity of the target fill light from a preset number of fill lights based on the image of the vehicle without fill light, and to acquire fill light images of the vehicle according to the on / off state and light intensity of the target fill light, wherein each fill light image includes a bright area and a shadow area;

[0019] a fusion module, configured to fuse the fill-light images to obtain a fused image, wherein each steel coil on the truck exhibits at least a first shadow region and a second shadow region, wherein the first shadow region is located on the left side of the steel coil and the second shadow region is located on the right side of the steel coil;

[0020] A positioning module is used to calculate the initial fitting coordinates and tilt angle of each steel coil based on the non-filled light image and the fused image, and calculate the center coordinates of each steel coil based on the initial fitting coordinates and the steel coil tilt angle.

[0021] In a third aspect, the present disclosure further provides an electronic device comprising: a processor and a memory, wherein the memory stores machine-readable instructions executable by the processor, and the processor is configured to read the machine-readable instructions from the memory and execute the machine-readable instructions to implement the steel coil positioning method based on a pallet truck provided in an embodiment of the present disclosure.

[0022] In a fourth aspect, the present disclosure further provides a computer-readable storage medium storing a computer program, wherein the computer program is used to execute the steel coil positioning method based on a pallet truck provided in an embodiment of the present disclosure.

[0023] In summary, the present disclosure provides a method, device, equipment and medium for positioning steel coils based on a pallet truck. By parking the pallet truck transporting steel coils in a designated area within a parking space, it ensures that the camera above the parking space can capture an image covering the entire pallet truck. Furthermore, the pallet truck's unfilled light image and the filled light image after the pallet truck is filled with light in multiple different directions are collected. By performing a depth analysis of the unfilled light image, each fill light is reasonably turned on and the light intensity value of each target fill light is dynamically adjusted, effectively reducing the influence of ambient light on the camera's fill light imaging and improving the accuracy of the collected filled light image. Moreover, each filled light image is in a bright area and a shadow area, which makes each steel coil present multiple shadow areas in the fused image obtained after the fill light images from different directions are fused. By analyzing and calculating the position, shape, size and other information of the shadow area on the steel coil, the center of the steel coil is obtained, realizing the joint solution of multi-directional fill light, and at the same time estimating the pitch angle and roll angle of the tilted steel coil, accurately correcting the center coordinates of the steel coil, and realizing fast and accurate positioning of the steel coil. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solutions in the present disclosure or the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0025] Figure 1 This is a schematic diagram of a process for positioning a steel coil based on a pallet truck provided by the present disclosure;

[0026] Figure 2a This is a schematic diagram of a parking space provided by the present invention from a bird's-eye view;

[0027] Figure 2b This is a schematic diagram of a parking space provided by the present invention from a frontal perspective;

[0028] Figure 2c is a schematic diagram of a shaded area of ​​a steel coil on a pallet truck provided by the present disclosure;

[0029] Figure 3 1 is a flow chart of a method for calculating the center coordinates of each steel coil provided by the present disclosure;

[0030] Figure 4 This is a structural schematic diagram of a steel coil positioning device based on a cart provided by the present disclosure;

[0031] Figure 5 It is a structural diagram of an electronic device provided by an embodiment of the present disclosure. DETAILED DESCRIPTION

[0032] To make the objectives, technical solutions, and advantages of this disclosure more clear, the technical solutions of this disclosure will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are only part of the embodiments of this disclosure, not all of them. All other embodiments obtained by persons of ordinary skill in the art based on the embodiments of this disclosure without creative effort shall fall within the scope of protection of this disclosure.

[0033] Figure 1 This is a schematic diagram of a process flow of a steel coil positioning method based on a plate car provided by the present disclosure; Figure 1 , the method comprising:

[0034] Step S11 , collecting a non-fill light image of a cart, wherein the cart is located in a parking space and a horizontally placed steel coil is mounted on the cart, and a preset number of fill lights are installed in different directions above the parking space.

[0035] Specifically, it can be understood that the flatbed truck is a vehicle for transporting steel coils, and the steel coils are all placed horizontally on the flatbed truck, with the axis of the steel coil parallel to the loading plane and perpendicular to the side of the vehicle. The steel coils can be fixed on the flatbed truck by a saddle fixation method, a wedge fixation method, or other fixing methods. The flatbed truck generally transports the steel coils to the steel coil warehouse for unloading. The steel coil warehouse usually has a fixed area for parking the flatbed truck to facilitate loading or unloading of steel coils, that is, the parking space is set in the steel coil warehouse, and the parking space is a designated parking area for the flatbed truck. The parking space is usually a rectangular area, and the length and width of the rectangle can be determined according to the commonly used flatbed truck specifications. For example, the length and width of the rectangle are both greater than the length and width of commonly used flatbed trucks to ensure that the flatbed truck can fully enter the parking space.

[0036] In some embodiments, the cart being located in a parking space means that the cart is parked at a designated position in the parking space. The designated position may be a fixed line in the parking space, for example, the wide side of the parking space or a line segment in the parking space. Taking the wide side of the parking space as an example, it is necessary to align the front end of the cart with the wide side of the parking space and the midpoint of the front end coincides with the midpoint of the wide side of the parking space when parking the cart. The designated position may also be at least one fixed point in the parking space, for example, a vertex of the parking space or a fixed point in the parking space. Taking a vertex of the parking space as an example, it is necessary to park a corner of the cart at the vertex when parking the cart.

[0037] Figure 2a It is a schematic diagram of a parking space provided by the present invention from a top-down angle. The parking space 210 is a rectangular area surrounded by four outer sides AB, BC, CD and DA. The cart 220 loaded with steel coils 230 is parked in the middle of the parking space 210 without any tilt. That is, the length direction of the cart 220 is parallel to the upper side DC or the lower side AB of the parking space 210, the width direction coincides with the right side BC of the parking space 210, and the midpoint of the front end of the vehicle coincides with the midpoint of the right side BC.

[0038] In some embodiments, at least one indicator light may be provided around the parking space. The at least one indicator light is used to detect and prompt the driver of the cart whether the cart is parked at a designated position within the parking space. For example, ultrasonic ranging or other algorithms may be used for detection. When the cart is parked at a designated position within the parking space, the indicator light is green; when the cart is not parked at a designated position within the parking space, the indicator light is red.

[0039] Specifically, it can also be understood that the image without fill light refers to the image of the cart captured by the camera when the cart is in the parking space, that is, the image without fill light is the image of the cart captured after the cart has stopped, and is generally used to analyze the ambient light in the steel coil warehouse. The camera is installed directly above the center of the parking space, and is used to scan and image the cart in the parking space. For example, for parking space 210, the center of the parking space is the intersection of AC and BD. Installing the camera directly above the center of the parking space means installing the camera directly above the intersection of AC and BD, thereby ensuring that the camera takes a bird's-eye view of the cart in the parking space and the bird's-eye view image can cover the entire parking space area. The camera can be a high-resolution industrial camera with characteristics such as high frame rate and wide dynamic range, which can adapt to image acquisition under different lighting conditions. Since the saddle, wedge or other objects that fix the steel coil on the cart may change or even break as the cart moves during transportation, this will indirectly cause the steel coil to tilt at multiple angles and in multiple directions. For example, the movement of the wedge causes the steel coil to tilt to the left or right or to the front and back. Therefore, the multi-directional tilt of the steel coil can be calculated by comparing and analyzing the fill light in different directions. Therefore, in order to achieve multi-directional fill light, fill lights in multiple directions are required. A preset number of fill lights are installed in different directions above the parking space. The fill lights are used to fill light the steel coil on the cart from different directions to ensure that the fill light in each direction can clearly outline the three-dimensional contour of the edge of the steel coil, thereby obtaining clearer and more accurate steel coil contour information. Each fill light is located at the same height, and each fill light and camera can be located at the same height or at different heights. This is not specifically limited in the present disclosure.

[0040] In some embodiments, the spatial coordinate system and the image coordinate system are established with the rectangular area corresponding to the parking space as the reference benchmark, see Figure 2a , a spatial coordinate system is constructed with the vertex A at the lower left corner of the parking space 210 as the origin, the extending direction of the lower side AB of the parking space 210 as the X-axis, and the extending direction of the left side AD of the parking space 210 as the Y-axis. Similarly, in the acquired image without fill light, an image coordinate system is constructed with the vertex A at the lower left corner of the parking space as the origin, the long side of the parking space as the X-axis, and the wide side of the parking space as the Y-axis. Since the origin of the spatial coordinate system and the image coordinate system is the same, the conversion relationship between the spatial coordinate system and the image coordinate system can be determined based on a specific length and its pixel value in the image without fill light captured by the camera. For example, a certain point in the image coordinate system ( x i , y i ) corresponds to the point in the space coordinate system ( X i , Y i )=( kxi , ky i ),in, k It can be the ratio of the actual length of the parking space to the pixel value of the long side of the parking space in the image without fill light, or it can be the ratio of the actual width of the parking space to the pixel value of the wide side of the parking space in the image without fill light.

[0041] In some embodiments, a preset number of fill lights are installed in different directions above the parking space, including: at least one fill light is installed in at least four directions of front, rear, left and right above the parking space, wherein the front and rear directions above the parking space correspond to the Y-axis direction of the spatial coordinate system, and the left and right directions above the parking space correspond to the X-axis direction of the spatial coordinate system. In addition to the four directions of front, rear, left and right, fill lights can also be installed in the left front, left rear, right front, right rear and other directions, and the present disclosure does not limit this.

[0042] In some embodiments, the preset number can be set to 4, or other values ​​greater than 4, and this disclosure does not impose any particular limitation thereto. When the preset number is 4, a fill light can be installed in each of the four directions above the parking space: front, rear, left, and right. For example, the fill light installed in the front direction can be directly above the midpoint of the upper edge of the parking space, the fill light installed in the rear direction can be directly above the midpoint of the lower edge of the parking space, the fill light installed in the left direction can be directly above the midpoint of the left side of the parking space, and the fill light installed in the right direction can be directly above the midpoint of the right side of the parking space. When the preset number is 8, one fill light may be installed in each of the eight directions of front, rear, left, right, left front, left rear, right front, and right rear above the parking space, or two fill lights may be installed in each of the four directions of front, rear, left, and right above the parking space. For example, the two fill lights in the front direction may be installed directly above the upper 1 / 3 and 2 / 3 of the parking space, the two fill lights in the rear direction may be installed directly above the lower 1 / 3 and 2 / 3 of the parking space, the two fill lights in the left direction may be installed directly above the left 1 / 3 and 2 / 3 of the parking space, and the two fill lights in the right direction may be installed directly above the right 1 / 3 and 2 / 3 of the parking space.

[0043] It should be noted that a preset number of fill lights are symmetrically distributed relative to the camera, and the illumination angle of each fill light is adjustable and can satisfy the requirement that a shadow area exists on each steel coil after illumination. For example, with the camera as the center, the fill lights in the front direction and the fill lights in the rear direction are symmetrically installed, and the fill lights in the left direction and the fill lights in the right direction are symmetrically installed. Moreover, the distance between the fill lights and the camera can be adjusted according to actual needs. They can be directly above the midpoint of the edge of the parking space, or directly above a point within the boundary of the parking space and symmetrically distributed with the camera as the center to meet normal illumination. This disclosure does not make any special restrictions on this.

[0044] Figure 2b 2 is a schematic diagram of a parking space provided by the present disclosure from a frontal perspective. A camera 240 and two fill lights, one on the left and one on the right, are mounted in the center above the parking space 210. Specifically, a fill light 250 is mounted on the left side of the parking space 210, and a fill light 260 is mounted on the right side of the parking space 210. The fill light 250 is located to the left of the four steel coils, and the fill light 260 is located to the right of the four steel coils. The distance between the fill light 250 and the camera 240 is equal to the distance between the fill light 260 and the camera 240. In some embodiments, the fill lights 250, 260, and the camera 240 are mounted on the same slide rail. The camera 240 is fixed, while the fill lights 250 and 260 can be fixed or slidable. The length of the slide rail covers the length of the parking space, allowing the left and right fill lights to move on the slide rail, achieving more flexible fill lighting.

[0045] In the above method, by parking the truck transporting the steel coils in a designated area within the parking space, it is ensured that the camera above the parking space can capture an image covering the entire truck for positioning the steel coils on the truck; by arranging fill lights in multiple directions above the parking space, the fill lights in each direction can be used to clearly outline the edge of the steel coil.

[0046] Step S12, based on the non-fill light image of the cart, determining the target fill light and the light intensity of the target fill light from a preset number of fill lights, and collecting each fill light image of the cart according to the switch and light intensity of the target fill light, each fill light image including a bright area and a shadow area.

[0047] The preset number of fill lights can be ordinary monochromatic LED light sources, with an initial illumination angle of 30 degrees. In some embodiments, since indoor ambient light primarily consists of sunlight and lamplight, to better filter out ambient light, the fill lights can also be monochromatic LED light sources equipped with narrowband filters. In this case, the camera is equipped with narrowband filters of corresponding wavelengths to fully filter out other wavelength components of the ambient light, such as visible light.

[0048] Specifically, it is understood that, for the steel coils lying on the cart, if the steel coils have a pitch tilt along the Y-axis, the outline of the horizontal steel coils will appear non-rectangular in the unfilled light image taken from a top-down perspective; if there is no pitch tilt along the Y-axis, regardless of whether there is a rolling tilt along the X-axis, the outline of the horizontal steel coils will generally appear rectangular. To this end, a preliminary identification and judgment of each steel coil on the cart is performed based on the unfilled light image. If the steel coils clearly do not have a pitch tilt along the Y-axis, then when capturing the fill-light image, only fill-light images in at least the left and right directions need to be acquired. That is, only the fill-lights in the left and right directions need to be determined as the target fill-lights to be turned on.

[0049] In some embodiments, determining the target fill light and the light intensity of the target fill light from a preset number of fill lights based on the image of the vehicle without fill light specifically includes the following steps:

[0050] Step S121, performing edge detection on the non-fill light image of the cart to determine the contour shape of each steel coil. If the contour shape of each steel coil is a rectangle, the target fill light includes at least fill lights in the left and right directions; otherwise, the target fill light includes at least fill lights in the left, right, front and rear directions.

[0051] Specifically, the acquired image without fill light is first preprocessed, for example, the image without fill light is grayscaled to reduce the amount of image data, the image without fill light is contrast enhanced to enhance the contrast between the steel coil and the background in the image, the image without fill light is filtered using a median filter algorithm to remove noise, etc.; then, the preprocessed image without fill light is edge detected to extract the edge contour of the steel coil in the image, for example, the image without fill light of the cart is edge detected using a target detection algorithm, the contour information of each steel coil on the cart is identified, and the discontinuous edges are connected into a complete steel coil contour through a contour closing operation, thereby determining the contour shape of each steel coil. The target detection algorithm can be a canny edge detection algorithm or other detection algorithms such as YOLO, and the present disclosure does not limit this.

[0052] More specifically, if the outline shape of each steel coil is a rectangle, it indicates that the steel coil does not have any pitch tilt along the Y-axis direction. At this time, it is only necessary to turn on the fill lights in the left and right directions for fill light; otherwise, if the outline shape of any steel coil is non-rectangular, it indicates that the steel coil must have a pitch tilt along the Y-axis direction and may have a rolling tilt along the X-axis direction. At this time, it is necessary to turn on the fill lights in at least four directions of front, back, left and right for fill light.

[0053] Step S122: Calculate the intensity of the ambient light based on the image without fill light.

[0054] Specifically, the ambient light includes sunlight, indoor lighting, etc. The intensity of the ambient light refers to the brightness change of the ambient light on the steel coil, and the intensity of the ambient light includes the global contrast of each steel coil.

[0055] In some embodiments, for each coil in the non-filled light image, the intensity of the ambient light is represented by the global contrast of each coil. For example, the corresponding target area is determined based on the contour shape of each coil, and the global contrast G of each coil in the target area is calculated as (L max -L min ) / (L max +L min ), where L max Indicates the maximum brightness value within the target area corresponding to each steel coil, L min Indicates the minimum brightness value within the target area corresponding to each steel coil.

[0056] Step S123 : if the intensity of the ambient light meets the preset requirement, the target light intensity value of each target fill light is adaptively calculated according to the image without fill light; otherwise, the light intensity of the target fill light is set to the preset light intensity value.

[0057] Specifically, the preset requirements include that the global contrast of at least one steel coil is greater than a first threshold value, and the intensity of the ambient light meets the preset requirements, that is, the global contrast of at least one steel coil calculated is greater than the first threshold value. At this time, it is considered that the ambient light has a significant impact on the brightness of any steel coil, and it is necessary to readjust the light intensity value of the target fill light according to the actual ambient light to reduce the impact of the ambient light on the fill light. Otherwise, the intensity of the ambient light does not meet the preset requirements, that is, the global contrast of each steel coil calculated is less than or equal to the first threshold value. At this time, it is considered that the ambient light has a small impact on the brightness of all steel coils, and the light intensity value of the target fill light can be directly set according to the preset light intensity value. Among them, the first threshold is set according to the empirical value, and the present disclosure does not limit this.

[0058] In some embodiments, the preset light intensity value can be determined based on the image without fill light. For example, for each coil in the image without fill light, the average brightness value of each coil is first calculated based on the contour shape of each coil determined in step S121, and then the average brightness value of each coil is calculated to obtain a comprehensive brightness value for all coils. Then, based on the comprehensive brightness value, the preset light intensity value is determined. For example, the preset light intensity value is three or five times the initial light intensity value, thereby ensuring that the brightness of the fill light area is significantly higher than the ambient light, thereby preventing shadows from being submerged by the ambient light during fill light.

[0059] In other embodiments, after determining multiple target fill lights, if the ambient light has a significant impact on the steel coils on the cart, it is necessary to turn on each target fill light in sequence and dynamically adjust the light intensity value emitted by each target fill light according to the actual fill light situation. The adaptive calculation of the target light intensity value of each target fill light based on the image without fill light specifically includes:

[0060] Step a1: for any target fill light, determining the initial light intensity of the target fill light based on the image without fill light;

[0061] Specifically, the same method as the above-mentioned "determining the preset light intensity value based on the image without fill light" is adopted. For each steel coil in the image without fill light, the average brightness value of each steel coil is first calculated based on the contour shape of each steel coil determined in step S121, and then the average of each steel coil is calculated to obtain the comprehensive brightness value of all steel coils. Then, the preset light intensity value is determined based on the comprehensive brightness value. For example, the preset light intensity value is three times the initial light intensity value, or the preset light intensity value is five times the initial light intensity value, or other multiples.

[0062] Step a2, gradually adjusting the light intensity based on the initial light intensity and sequentially acquiring fill light images under various light intensities until the fill light image meets the fill light requirement, and stopping adjusting the light intensity, and determining the light intensity that meets the fill light requirement as the target light intensity value of the target fill light.

[0063] Specifically, if the target fill light is a left-facing fill light, a shadow area will be present on the right side of each steel coil in the fill light image; if the target fill light is a right-facing fill light, a shadow area will be present on the left side of each steel coil in the fill light image. The fill light image meets the fill light requirements if the shadow area of ​​the fill light image is relatively prominent relative to each steel coil or the entire fill light image.

[0064] More specifically, because the binding tape on a steel coil is relatively dim relative to the coil itself and can be easily mistaken for its outline, and the coil's binding tape distribution is irregular, capturing the image from each directional fill light requires careful analysis of the actual fill light situation for each coil. This means adjusting the fill light intensity for each individual fill light individually. For each target fill light, after determining the initial light intensity based on the unlit image, the target fill light is first turned on and illuminates the cart at the initial intensity. The camera is then controlled to capture the fill light image at this time. The fill light image is grayscaled, filtered, and processed to calculate the average grayscale value G1. Object detection is then used to extract the edge contours of the coils in the fill light image, thereby determining the area corresponding to each coil and the shadow region corresponding to each coil. The grayscale mean G2 of the shadow region within each coil's corresponding area is then calculated, and the grayscale difference ΔG = G1 - G2 for each coil is further calculated. Finally, whether the shadow area of ​​each steel coil is obvious is determined by judging the size of the grayscale difference corresponding to each steel coil and the second threshold. When the grayscale difference ΔG of all steel coils is greater than or equal to the second threshold, it indicates that the shadow areas of all steel coils are relatively obvious under the light intensity of the current fill light. At this time, the light intensity value can be directly used as the light intensity value of the target fill light; when among all the steel coils, there is a steel coil with a grayscale difference ΔG less than the second threshold, which indicates that the shadow area of ​​the steel coil is relatively unclear. At this time, it is necessary to increase the light intensity value of the target fill light for fill light enhancement, and re-capture the fill light image to recalculate the grayscale difference and the corresponding threshold judgment until the grayscale difference ΔG of all steel coils is greater than or equal to the second threshold.

[0065] In the above embodiment, the stepwise adjustment of the light intensity based on the initial light intensity can be achieved by setting a light intensity increment. For example, the initial light intensity value is P0, the light intensity increment is ΔP, and the next light intensity value is calculated sequentially by the light intensity increment of ΔP, such as the light intensity value from P0 to ΔP+P0 and 2ΔP+P0. The stepwise adjustment of the light intensity based on the initial light intensity can also be achieved by setting a gain factor. For example, the initial light intensity value is P0, the gain factor is w, and the next light intensity value is calculated sequentially by the gain factor w, such as the light intensity value from P0 to w×P0. This disclosure is not limited to this.

[0066] Specifically, it can also be understood that after determining the target fill light, each target fill light is turned on in turn and the camera is controlled to collect corresponding fill light images. Only one fill light is turned on at a time and the remaining fill lights remain off. Each fill light image includes a bright area and a shadow area. For example, when the left-direction fill light is turned on, the left side of each steel coil on the cart is a bright area and the right side is a shadow area. When the right-direction fill light is turned on, the right side of each steel coil on the cart is a bright area and the left side is a shadow area.

[0067] In some embodiments, if the target fill light includes a left-direction fill light and a right-direction fill light, the light intensities of the corresponding target fill lights include: a first light intensity of the left-direction fill light and a second light intensity of the right-direction fill light. Capturing fill light images of the cart based on the on / off and light intensities of the target fill lights includes: turning on the left-direction fill light and irradiating the cart with light of a first light intensity while keeping the right-direction fill light off, controlling the camera to capture a first fill light image, in which a shadow area appears on the right side of each steel coil; turning on the right-direction fill light and irradiating the cart with light of a second light intensity while keeping the left-direction fill light off, controlling the camera to capture a second fill light image, in which a shadow area appears on the left side of each steel coil.

[0068] In the above method, the inclination of the steel coil is preliminarily analyzed based on the contour of the steel coil in the image without fill light so as to reasonably turn on the corresponding fill light for fill light; the influence of the actual ambient light on the steel coil imaging is analyzed based on the global contrast of the image without fill light, and then the light intensity of each fill light is determined according to the influence of the ambient light, thereby reducing the influence of the ambient light and ensuring that the fill light image of each steel coil can clearly outline the edge contour; the target light intensity value of the final fill light is determined based on the brightness change of the steel coil area in each fill light image, so as to avoid problems such as unclear fill light imaging of the steel coil caused by ambient light and the straps on the steel coil.

[0069] Step S13, fusing the fill-light images to obtain a fused image, in which each steel coil on the pallet truck shows at least a first shadow area and a second shadow area, wherein the first shadow area is located on the left side of the steel coil, and the second shadow area is located on the right side of the steel coil.

[0070] Specifically, it can be understood that the camera successively captures multiple fill-light images with fill-light in different directions at the same angle at the corresponding position, and each fill-light image includes a bright area and a shadow area. In order to display all the shadow areas in each fill-light image in the same image, multiple fill-light images need to be fused.

[0071] In some embodiments, the fusion of the fill-in light images to obtain the fused image may adopt a mask-based fusion method. The fusion of the fill-in light images to obtain the fused image specifically includes:

[0072] Step S131 , constructing a mask for each fill-light image, wherein the position of the mask corresponding to the shadow area of ​​the fill-light image is 1, and the remaining positions are 0;

[0073] Specifically, edge detection is performed on each fill-in-light image to identify the contour information of each steel coil and the position of the shadow area, and then a mask is constructed for the fill-in-light image according to the position of the shadow area. The size of the mask is the same as the image size, and the mask is set to position 1 corresponding to the shadow area in the fill-in-light image, and the remaining positions are 0.

[0074] Step S132: multiplying each fill-in light image by its own mask to obtain a shadow image corresponding to each fill-in light image;

[0075] Specifically, a corresponding mask is constructed for each fill-in-light image based on the position of the shadow area on the steel coil. By multiplying the fill-in-light image and the mask, an image of only the shadow area can be obtained. That is, the shadow area on the steel coil can be completely preserved in each fill-in-light image.

[0076] Step S133 : superimposing the shadow images corresponding to the fill-in light images and any fill-in light image to obtain a fused image.

[0077] Specifically, by superimposing the multiple shadow images obtained in step S132 , an image including the shadow areas under all fill-lights can be obtained, and then superimposing the image with any fill-light image can obtain a complete fused image.

[0078] Figure 2c is a schematic diagram of the shadow area of ​​a steel coil on a pallet truck provided by the present disclosure. Figure 2c There is no obvious Y-axis tilt in each coil, and Figure 2c The shaded area is Figure 2b Take the left and right fill lights in the example as an example to perform fill light and image acquisition to obtain the fused image. Figure 2c When the left-direction fill light is turned on, the right-side shadows of the four steel coils on the cart are obtained, and when the right-direction fill light is turned on, the left-side shadows of the four steel coils on the cart are obtained. After fusion, each steel coil appears as a left shadow and a right shadow.

[0079] In other embodiments, the fusion of the fill-in images to obtain a fused image may also be performed using a frequency domain decomposition fusion method. The fusion of the fill-in images to obtain a fused image may specifically include: performing image decomposition on each fill-in image to obtain decomposed images, wherein the decomposed images include at least a low-frequency layer and a high-frequency layer, for example, performing Laplacian pyramid decomposition on each fill-in image to obtain a low-frequency layer and a high-frequency layer; performing weighted fusion on each layer and then performing an inverse transform to generate a fused image, wherein the weighted fusion of the low-frequency layer can preserve low-light information in shadow areas, and the high-frequency layer can fuse details from multiple images and preserve shadow details in all images.

[0080] In the above method, a fused image is obtained by fusing multiple fill-light images, and the shadow information obtained by multi-directional fill-light is presented in the fused image, effectively retaining all shadow features of the steel coil.

[0081] Step S14: calculating the initial fitting coordinates and the tilt angle of each steel coil based on the non-filled light image and the fused image, and calculating the center coordinates of each steel coil based on the initial fitting coordinates and the steel coil tilt angle.

[0082] Specifically, it can be understood that the initial fitting coordinates are determined based on the contour information of the steel coil or based on the shadow information of the steel coil. First, edge detection is performed on each steel coil on the pallet based on the non-filled light image to extract the contour information of each steel coil and determine the contour shape of the steel coil. The plane coordinates of each steel coil in the non-filled light image currently collected from a bird's-eye view are further determined as the initial fitting coordinates. Then, based on the fused image obtained by fusing multiple fill-light images, the multiple shadow areas presented by each steel coil are analyzed to calculate the inclination angle of the steel coil in the left-right direction and / or the front-back direction. Finally, based on the initial fitting coordinates and the inclination angle, the actual coordinates of the steel coil after the inclination are calculated.

[0083] In the above method, by parking the flatbed truck transporting steel coils in a designated area within the parking space, the camera above the parking space is able to capture an image covering the entire truck. Furthermore, an unlit image of the truck and a lit image of the truck after multiple directional fill-lighting are collected. Through in-depth analysis of the unlit image, each fill-light is properly activated and the light intensity of each target fill-light is dynamically adjusted, effectively reducing the impact of ambient light on the camera's fill-light imaging and improving the accuracy of the collected lit image. Furthermore, each fill-light image is in both bright and shadow areas, resulting in multiple shadow areas for each steel coil in the fused image obtained by fusing the fill-light images from different directions. By analyzing and calculating information such as the position, shape, and size of the shadow areas on the steel coil, the center of the steel coil is determined, achieving a joint solution for multi-directional fill-lighting. Simultaneously, the pitch and roll angles of the tilted steel coil are estimated, and the center coordinates of the steel coil are accurately corrected, achieving rapid and precise positioning of the steel coil.

[0084] Figure 3 This is a flow chart of a method for calculating the center coordinates of each steel coil provided by the present disclosure; Figure 3 , the calculating the initial fitting coordinates and the tilt angle of each steel coil based on the non-filled light image and the fused image, and calculating the center coordinates of each steel coil based on the initial fitting coordinates and the steel coil tilt angle, including:

[0085] Step S31 : calculating the initial fitting coordinates of each steel coil on the cart based on the fused image and the image without fill light.

[0086] Specifically, when calculating the initial fitting coordinates, the initial fitting coordinates may be determined based on the contour information of the steel coil or based on the shadow information of the steel coil.

[0087] In some embodiments, the calculation of the initial fitting coordinates can be based on the non-filled light image or any fill-light image, and the contour information of each steel coil can be extracted by performing edge detection on the image and then roughly calculated. Taking the non-filled light image as an example, it is necessary to perform grayscale conversion, filtering, target detection and other processing on the non-filled light image to extract the contour information of each steel coil, and further determine the contour shape of each steel coil. If the contour shape of the steel coil is a rectangle, the initial fitting coordinates can be obtained based on the X-coordinate and Y-coordinate of the center of the rectangle. If the contour shape of the steel coil is non-rectangular, such as a parallelogram or other irregular quadrilateral, the X-axis coordinate of the initial fitting coordinates can be calculated based on the left and right contour boundaries of each steel coil, and the Y-axis coordinate of the initial fitting coordinates can be calculated based on the upper and lower contour boundaries of each steel coil.

[0088] In other embodiments, the shadow of the steel coil in the fused image can clearly outline the three-dimensional contour of the steel coil edge compared to the contour determined by edge detection based on the image without fill light. Therefore, the calculation of the initial fitting coordinates can also be performed based on the shadow information of the fused image. In this case, the calculation of the initial fitting coordinates of each steel coil on the cart based on the fused image and the image without fill light includes at least:

[0089] If the shadow area of ​​any steel coil is located on the left and right sides of the steel coil, the X-axis coordinate of the initial fitting coordinate is determined according to the shadow area, and the Y-axis coordinate of the initial fitting coordinate is determined according to the upper and lower contours of the steel coil. The Y-axis coordinate of the initial fitting coordinate determined according to the shadow area can be determined by first detecting the upper and lower contours of the steel coil and then determining the Y-axis coordinate y0 according to the upper and lower contours; the X-axis coordinate of the initial fitting coordinate determined according to the shadow area includes: determining the X-axis coordinate of the right boundary of the left shadow area according to the left shadow area as x1, and the width of the left shadow area as W1; determining the X-axis coordinate of the left boundary of the right shadow area according to the right shadow area as x2, and the width of the right shadow area as W1, then the calculation formula of the X-axis coordinate x0 of the initial fitting coordinate is: .

[0090] If the shadow area of ​​any steel coil is located on the upper or lower side of the steel coil and the shadow area is an ellipse, the X-axis coordinate of the initial fitting coordinate is determined according to the left and right contours of the steel coil, and the Y-axis coordinate of the initial fitting coordinate is calculated according to the minor semi-axis of the ellipse corresponding to the shadow area. For example, the Y-axis coordinate of the initial fitting coordinate is calculated according to the Y-axis coordinate of the lower boundary of the shadow area and the Y-axis coordinate of the lower side contour of the steel coil. For another example, the Y-axis coordinate of the initial fitting coordinate is calculated according to the Y-axis coordinate of the upper boundary of the shadow area and the Y-axis coordinate of the upper side contour of the steel coil.

[0091] Step S32: determining feature information of the shadow area in each steel coil based on the fused image, and calculating the tilt angle of each steel coil based on the feature information of the shadow area in each steel coil.

[0092] Specifically, it can be understood that the characteristic information of the shadow area includes the position, shape, and size of the shadow area. The position of the shadow area refers to whether the shadow area is located above, below, left, or right within the outline of the steel coil. The shape of the shadow area is generally divided into rectangular, trapezoidal, and elliptical according to the actual shadow generated. When the steel coil does not have a pitch angle in the front-to-back direction, regardless of whether the steel coil has a roll angle in the left-to-right direction, the shadow area of ​​the steel coil in the fill-light image taken by the camera from overhead is rectangular. When the steel coil has a pitch angle in the front-to-back direction, the shadow area of ​​the steel coil in the fill-light image taken by the camera from overhead after fill-lighting in the front-to-back direction may include an elliptical shadow area. The size of the shadow area refers to the size of the shadow area, for example, the length or width of a rectangular shadow, the major semi-axis or minor semi-axis of an elliptical shadow, the upper base, lower base, or waist length of a trapezoidal shadow, etc.

[0093] Specifically, it can also be understood that after determining the position, shape, size and other characteristic information of the shadow area in each steel coil based on the fused image, the inclination angle of each steel coil is further calculated based on the characteristic information of these shadow areas.

[0094] In some embodiments, the tilt angle includes: a roll angle and a pitch angle. The calculating the tilt angle of each steel coil based on the characteristic information of the shadow area in each steel coil at least includes:

[0095] b1. If the shadow area of ​​any steel coil is located on the left and right sides of the steel coil and the shadow area is rectangular, the tilt angle of the steel coil is calculated according to the width of the rectangle corresponding to the shadow area;

[0096] Specifically, when the shadow area of ​​the steel coil is located on the left and right sides of the steel coil, the tilt angle of the steel coil should be the rolling angle in the left and right directions. In this case, for any steel coil that presents a first shadow area and a second shadow area, the first shadow area is located on the left side of the steel coil, the second shadow area is located on the right side of the steel coil, and both the first shadow area and the second shadow area are rectangular, then the width of the first shadow area and the width of the second shadow area are calculated first, and then the tilt angle of the steel coil - the rolling angle - is calculated based on the first formula. The first formula is:

[0097]

[0098] in, θ1 is the rolling angle of the steel coil, W1 is the width of the rectangle corresponding to the first shaded area, W2 is the width of the rectangle corresponding to the second shaded area, and R is the actual radius of the steel coil, which is generally obtained by querying the transportation information of the flatbed truck.

[0099] b2. If the shadow area of ​​any steel coil is located on the upper side or the lower side of the steel coil and the shadow area is an ellipse, the inclination angle of the steel coil is calculated according to the minor semi-axis of the ellipse corresponding to the shadow area.

[0100] Specifically, when the shadow area of ​​the steel coil is located on the upper or lower side of the steel coil, the inclination angle of the steel coil should be the pitch angle in the front-to-back direction. It is worth noting that after the steel coil tilts in the front-to-back direction, the fill-light image collected after fill-lighting with the front direction fill-light or the rear direction fill-light has a shadow area only on one side. For example, if the inclination of the steel coil in the front-to-back direction is that the front side is raised at a certain angle, then there is no shadow in the fill-light image of the front direction fill-light, and there will be an elliptical shadow in the upper half of the steel coil outline in the fill-light image of the rear direction fill-light; and if the inclination of the steel coil in the front-to-back direction is that the rear side is raised at a certain angle, then there will be an elliptical shadow in the lower half of the steel coil outline in the fill-light image of the front direction fill-light, and there will be no shadow in the fill-light image of the rear direction fill-light.

[0101] More specifically, if the shadow area of ​​any coil is located above or below the coil and is an ellipse, first calculate the minor semi-axis of the ellipse corresponding to the shadow area, and then calculate the inclination angle of the coil, the pitch angle, according to the second formula:

[0102]

[0103] in, θ 2 is the pitch angle of the steel coil, W3 is the short semi-axis of the ellipse corresponding to the shaded area, and L is the actual length of the steel coil, which is generally obtained by querying the transportation information of the truck.

[0104] Step S33: Calculate the center coordinates of each steel coil based on the initial fitting coordinates and the steel coil inclination angle.

[0105] Specifically, it can be understood that the tilt angle includes the roll angle and / or the pitch angle, that is, each steel coil may only have a tilt in the left-right direction, or may only have a tilt in the front-back direction, or may have a tilt in the left-right direction and a tilt in the front-back direction at the same time. Therefore, when the steel coil is tilted, the initial fitting coordinates need to be corrected accordingly. The tilt in the left-right direction requires the calculation of the left-right offset, and the tilt in the front-back direction requires the calculation of the front-back offset.

[0106] Specifically, it can also be understood that the calculation of the center coordinates of each steel coil based on the initial fitting coordinates and the steel coil inclination angle includes:

[0107] Step S331, calculating a first offset of the center coordinates of the steel coil based on the roll angle of each steel coil, and / or calculating a second offset of the center coordinates of the steel coil based on the pitch angle of each steel coil;

[0108] Specifically, the first offset is an offset in the left-right direction, that is, an X-axis offset; the second offset is an offset in the front-back direction, that is, a Y-axis offset.

[0109] The first offset of the center coordinate of the steel coil is calculated based on the rolling angle of each steel coil. The first offset ΔX can be calculated according to the third formula. The third formula is:

[0110]

[0111] The second offset of the center coordinate of the steel coil is calculated based on the pitch angle of each steel coil. The second offset ΔY can be calculated according to the fourth formula. The fourth formula is:

[0112]

[0113] Step S332: Calculate the center coordinates of each steel coil based on the initial fitting coordinates and the first offset and / or the second offset.

[0114] Specifically, assuming that the center coordinates of each steel coil are ( x , y ), then according to the initial coordinates ( x 0, y 0), the first offset ΔX and the second offset ΔY are used to calculate the center coordinates of each coil, that is, x = x 0+ΔX, y = y 0+ΔY.

[0115] It should be noted that the coordinate system corresponding to the center coordinate of the steel coil calculated in step S332 is the image coordinate system, and the center coordinate of the steel coil needs to be sent to the unmanned overhead crane system. Therefore, in actual application, the center coordinate of the steel coil needs to be converted from the center coordinate of the image coordinate system ( x , y ) is converted to the center coordinates of the spatial coordinate system ( X , Y )=( kx , ky In addition, the center coordinates of the steel coils determined in the embodiment of the present disclosure are the X-axis coordinates and the Y-axis coordinates. After the center coordinates of each steel coil are determined, the unmanned overhead crane system can be controlled to move the steel coils according to the center coordinates of the determined steel coils in the space coordinate system ( X , Y) is moved, and after moving to this point, the steel coil is clamped by lowering the clamp. During actual operation, the clamp lowering distance can be determined according to the clamping infrared sensor, or by the Z-axis coordinate of the steel coil in the spatial coordinate system. At this time, the Z-axis coordinate of the steel coil needs to be additionally calculated. The method for determining the clamp lowering distance is not specifically limited in this disclosure.

[0116] In this method, by processing and analyzing the fused image obtained by fusing the unlit image of the pallet truck and the lit images collected after the pallet truck was lit from multiple different directions, the position, shape, size, and other information of the shadow area on the steel coil can be accurately obtained. At the same time, by extracting the characteristic information such as the position, shape, and size of the shadow area in the fused image, the initial fitting coordinates of the steel coil center can be obtained, realizing the joint solution of multi-directional fill light. Then, through the precise analysis of the shadow area characteristics, the tilt angle of the steel coil under multi-directional fill light can be further calculated, thereby achieving the rapid and accurate positioning of all steel coils on the pallet truck.

[0117] Figure 4 This is a structural diagram of a steel coil positioning device based on a cart provided by the present disclosure. Figure 4 , the device 400 includes:

[0118] A first acquisition module 410 is configured to acquire an image of a cart without fill light, the cart being located in a parking space and loaded with horizontally placed steel coils, wherein a preset number of fill lights are installed in different directions above the parking space;

[0119] A second acquisition module 420 is configured to determine a target fill light and the light intensity of the target fill light from a preset number of fill lights based on the image of the vehicle without fill light, and to acquire fill light images of the vehicle according to the on / off state and light intensity of the target fill light, wherein each fill light image includes a bright area and a shadow area;

[0120] a fusion module 430 configured to fuse the fill-light images to generate a fused image, wherein each steel coil on the truck exhibits at least a first shadow region and a second shadow region, wherein the first shadow region is located on the left side of the steel coil and the second shadow region is located on the right side of the steel coil;

[0121] The positioning module 440 is configured to calculate the initial fitting coordinates and tilt angle of each steel coil based on the non-filled light image and the fused image, and calculate the center coordinates of each steel coil based on the initial fitting coordinates and the steel coil tilt angle.

[0122] For a detailed description of the above-mentioned steel coil positioning device based on a pallet truck, please refer to the description of the relevant method steps in the above-mentioned embodiment, and the repeated parts will not be repeated. The embodiments of the steel coil positioning method based on a pallet truck and the steel coil positioning device based on a pallet truck described above are merely illustrative, and the "modules" and "units" used therein as separate components may be a combination of software and / or hardware that implements a predetermined function, and may or may not be physically separated. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. A person of ordinary skill in the art can understand and implement it without any creative effort.

[0123] Figure 5 It is a structural diagram of an electronic device provided in an embodiment of the present disclosure. The electronic device 500 includes: a processor 510 and a memory 520. The processor 510 and the memory 520 are connected via a bus 530. The memory 520 stores machine-readable instructions executable by the processor 510. The processor 510 reads the machine-readable instructions from the memory 520 and executes the machine-readable instructions to implement the steel coil positioning method provided in an embodiment of the present disclosure.

[0124] It should be understood that the electronic device may be a personal computer (PC), a tablet computer, a smart phone, or other electronic device with logic computing capabilities.

[0125] The embodiment of the present disclosure further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program is used to execute the steel coil positioning method provided by the embodiment of the present disclosure.

[0126] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present disclosure, rather than to limit them. Although the present disclosure has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present disclosure, and should be covered by the protection scope of the present invention.

Claims

1. A steel coil positioning method based on a cart, characterized in that: The method comprises: Capturing an image of a cart without fill light, the cart being located in a parking space and carrying a horizontally placed steel coil, wherein a preset number of fill lights are installed in different directions above the parking space; Determining a target fill light and the light intensity of the target fill light from a preset number of fill lights based on the image of the vehicle without fill light, and collecting fill light images of the vehicle according to the on / off state and light intensity of the target fill light, wherein each fill light image includes a bright area and a shadow area; fusing the fill-light images to obtain a fused image, wherein each steel coil on the truck exhibits at least a first shadow region and a second shadow region, wherein the first shadow region is located on the left side of the steel coil and the second shadow region is located on the right side of the steel coil; The initial fitting coordinates and the tilt angle of each steel coil are calculated based on the non-filled light image and the fused image, and the center coordinates of each steel coil are calculated based on the initial fitting coordinates and the steel coil tilt angle.

2. The steel coil positioning method based on a pallet truck according to claim 1, characterized in that: The calculating of the initial fitting coordinates and the tilt angle of each steel coil based on the non-filled light image and the fused image, and calculating the center coordinates of each steel coil based on the initial fitting coordinates and the steel coil tilt angle, includes: Calculating the initial fitting coordinates of each steel coil on the cart based on the fused image and the image without fill light; determining feature information of a shadow region in each steel coil based on the fused image, and calculating a tilt angle of each steel coil based on the feature information of the shadow region in each steel coil; The center coordinates of each steel coil are calculated based on the initial fitting coordinates and the steel coil inclination angle.

3. The steel coil positioning method based on a pallet truck according to claim 2, characterized in that: The characteristic information of the shadow area includes the position, shape, and size of the shadow area. The calculating the tilt angle of each steel coil based on the characteristic information of the shadow area in each steel coil at least includes: If the shadow area of ​​any steel coil is located on the left and right sides of the steel coil and the shadow area is a rectangle, the tilt angle of the steel coil is calculated according to the width of the rectangle corresponding to the shadow area; If the shadow area of ​​any steel coil is located on the upper side or the lower side of the steel coil and the shadow area is an ellipse, the tilt angle of the steel coil is calculated according to the minor semi-axis of the ellipse corresponding to the shadow area.

4. The steel coil positioning method based on a pallet truck according to claim 3, characterized in that: The tilt angle includes a roll angle and / or a pitch angle, and the calculating the center coordinates of each steel coil based on the initial fitting coordinates and the steel coil tilt angle includes: Calculating a first offset of the center coordinates of the steel coil based on the roll angle of each steel coil, and / or calculating a second offset of the center coordinates of the steel coil based on the pitch angle of each steel coil; The center coordinates of each steel coil are calculated based on the initial fitting coordinates and the first offset and / or the second offset.

5. The steel coil positioning method based on a pallet truck according to claim 1, characterized in that: Fusing the fill light images to obtain a fused image, including: Constructing a mask for each fill-light image, wherein the mask is 0 in the bright area of ​​the fill-light image and is 1 in the shadow area of ​​the fill-light image; Multiply each fill-in light image by its own mask to obtain a shadow image corresponding to each fill-in light image; The shadow images corresponding to the fill-in light images are superimposed on any fill-in light image to obtain a fused image.

6. The steel coil positioning method based on a pallet truck according to claim 1, characterized in that: The method of determining a target fill light and the light intensity of the target fill light from a preset number of fill lights based on the image of the vehicle without fill light comprises: Performing edge detection on the unlit image of the cart to determine the outline of each steel coil; if the outlines of each steel coil are rectangular, the target fill lights include at least left and right fill lights; otherwise, the target fill lights include at least left, right, front, and rear fill lights; Calculating the intensity of the ambient light based on the image without fill light; If the intensity of the ambient light is less than or equal to a first threshold, the light intensity of the target fill light is set to a preset light intensity value; otherwise, the target light intensity value of each target fill light is adaptively calculated according to the image without fill light.

7. The steel coil positioning method based on a pallet truck according to claim 6, characterized in that: The adaptively calculating the target light intensity value of each target fill light according to the image without fill light specifically includes: For any target fill light, determining the initial light intensity of the target fill light based on the image without fill light; The light intensity is gradually adjusted based on the initial light intensity and fill light images under various light intensities are sequentially acquired until the fill light image meets the fill light requirement, and the light intensity that meets the fill light requirement is determined as the target light intensity value of the target fill light.

8. A steel coil positioning device based on a cart, characterized in that: The device comprises: a first acquisition module for acquiring an image of a cart without fill light, the cart being located in a parking space and carrying a horizontally placed steel coil, wherein a preset number of fill lights are installed in different directions above the parking space; a second acquisition module, configured to determine a target fill light and the light intensity of the target fill light from a preset number of fill lights based on the image of the vehicle without fill light, and to acquire fill light images of the vehicle according to the on / off state and light intensity of the target fill light, wherein each fill light image includes a bright area and a shadow area; a fusion module, configured to fuse the fill-light images to obtain a fused image, wherein each steel coil on the truck exhibits at least a first shadow region and a second shadow region, wherein the first shadow region is located on the left side of the steel coil and the second shadow region is located on the right side of the steel coil; A positioning module is used to calculate the initial fitting coordinates and tilt angle of each steel coil based on the non-filled light image and the fused image, and calculate the center coordinates of each steel coil based on the initial fitting coordinates and the steel coil tilt angle.

9. An electronic device, characterized in that: The electronic device includes: a processor and a memory, wherein the memory stores machine-readable instructions executable by the processor, and the processor is used to read the machine-readable instructions from the memory and execute the machine-readable instructions to implement the steel coil positioning method based on the pallet truck as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and the computer program is used to execute the steel coil positioning method based on a plate car as described in any one of claims 1 to 7.

Images