Full-space accurate distance measuring device and distance measuring method based on machine vision

Abstract

The invention discloses a full-space accurate distance measuring device and method based on machine vision, and the method comprises the following steps: S1, setting a distance measuring device, and building a coordinate system in a reservoir area; s2, when the target article enters the reservoir area, the identification information of the target article is stored as attribute information, the distance measuring device identifies the target article, and an identification result of the target article is added to the attribute information; s3, after the crown block carries the target object to the stacking area, the three laser range finders carry out range finding on the target object, coordinates of the target object are calculated, and the coordinates are added to the attribute information; and S4, when the crown block moves the warehousing target article according to the moving instruction, the article contacted with the crown block is verified, the article is moved after verification succeeds, and the moved coordinates are updated in the attribute information of the target article. The information of the article is correspondingly updated in the moving process of the article, so that the article can be conveniently identified.

Description

Technical Field

[0001] The present invention relates to a distance measurement method, in particular to a full-space precise distance measurement device and a distance measurement method based on machine vision. Background Art

[0002] At present, most factories choose robots to carry goods, which improves the production efficiency of the factory. The existing technology generally uses monocular vision for object recognition. The monocular vision system mainly relies on a visual sensor for imaging. This system projects the three-dimensional objective world onto a two-dimensional image during the imaging process, resulting in the loss of depth information and inaccurate identification of objects. There will be many items of cargo owners placed in the warehouse. When the items are put into the warehouse, it is necessary to distinguish the cargo owners to whom the items belong, especially when the shapes of the items are similar, which increases the difficulty of identifying the items. The robot's recognition of the items often makes mistakes, resulting in a lower efficiency of the robot's handling of items. In the process of moving the items, it is inevitable that the items will be confused, which reduces the accuracy of the robot's recognition of the items. Summary of the invention

[0003] One of the purposes of the present invention is to provide a full-space precise distance measuring device based on machine vision, so as to provide a distance measuring device that can effectively measure distance.

[0004] The second purpose of the present invention is to provide a full-space precise distance measurement method based on machine vision to solve the problem of low accuracy of robot recognition of objects in the prior art.

[0005] One of the purposes of the present invention is achieved in that:

[0006] A full-space precise distance measurement device based on machine vision, comprising:

[0007] The first distance measuring unit is connected to the master control robot and the power supply, respectively, and is installed on the crossbeam of the overhead travelling crane, and includes a chassis, a first laser distance meter and a first controller, and is used to measure the horizontal distance between the crossbeam and a specific wall surface;

[0008] The second distance measuring unit is connected to the master control robot and the power supply respectively, and is installed on the mobile trolley of the overhead travelling crane, and includes a chassis, a second laser distance meter, and a second controller, and is used to measure the horizontal distance between the mobile trolley and the shielding object at the end of the beam;

[0009] The third distance measuring unit is connected to the master control robot and the power supply respectively, and is installed on the hook of the mobile vehicle, and includes a chassis, a camera, a third laser rangefinder, a third controller and a dual millimeter wave radar, and is used to measure the vertical distance between the hook and the beam;

[0010] The master control robot is connected to the storage unit, the power supply and the three distance measuring units respectively, and is used to control the three controllers to work, receive the distance measuring information sent by each controller, and measure the three-dimensional coordinate information of the target object according to the received distance measuring information;

[0011] A storage unit connected to the master control robot and used to store attribute information of the item; and

[0012] The power supply is respectively connected to the master control robot, the storage unit and the three distance measuring units to provide working power for each part.

[0013] Furthermore, the first laser rangefinder and the second laser rangefinder are respectively replaced by distance sensors.

[0014] The second object of the present invention is achieved in this way:

[0015] A full-space precise distance measurement method based on machine vision includes the following steps:

[0016] S1. Set up a distance measuring device, the first distance measuring unit is set on the beam, used to measure the distance between the beam and the wall, the second distance measuring unit is set on the trolley, used to measure the distance between the trolley and the end of the beam, and the third distance measuring unit is set on the hook. When the hook contacts the object, the height distance between the hook and the beam in the vertical direction is measured; set a reference point in the warehouse area, and establish a coordinate system with the reference point as the origin;

[0017] S2. When the target item enters the storage area, the identification information of the target item is used as the attribute information of the target item and stored in the storage unit. The master control robot controls the third ranging unit to identify the target item and adds the identification result to the attribute information of the target item, wherein the identification information is the name of the owner;

[0018] S3. After the overhead crane carries the target object to the storage area, the three laser rangefinders measure the distance and send it to the master control robot via the corresponding controller. The master control robot calculates the coordinates of the target object based on the distance measured by the three laser rangefinders, adds the coordinates of the target object to the attribute information of the target object, and stores it in the storage unit;

[0019] S4. After the target object enters the storage area, when the overhead crane moves the target object according to the moving instruction, the master control robot verifies the target object contacted by the hook, and after the target object moves, recalculates the coordinates of the target object, and updates the coordinates in the attribute information of the target object according to the coordinates after the movement.

[0020] Furthermore, in step S2, the master control robot controls the third distance measuring unit to identify the target object in the following specific manner:

[0021] S2a-1. The master control robot sends an object recognition instruction to the third controller, and the third controller controls the camera to obtain a video stream of the object, and inputs the video stream obtained by the camera into the object recognition model therein, identifies and locks the target object, and obtains the category of the object;

[0022] S2a-2. The third controller sends the video stream of the locked target object to the dual millimeter wave radar, and the dual millimeter wave radar outlines the locked object and measures the size of the object.

[0023] Furthermore, in step S4, the specific method of verifying the target object contacted by the hook is:

[0024] S4a-1. When the hook contacts the object, the three laser rangefinders measure the distance and send it to the master control robot through their respective corresponding controllers. The master control robot calculates the three-dimensional coordinates of the object contacted by the hook based on the distance measurement information; if the calculated three-dimensional coordinates are different from the coordinates in the movement instruction, the master control robot sends an adjustment instruction to the overhead traveling crane until the coordinates of the object contacted by the hook are the same as the coordinates in the movement instruction; if the coordinates of the object contacted by the hook are the same as the coordinates in the movement instruction, execute step S4a-2;

[0025] S4a-2. The third controller controls the camera to obtain the video stream of the object, and inputs the video stream to the object recognition model. The object recognition model identifies the type of object, and the third controller locks the identified object. Then the third controller controls the dual millimeter-wave radar to outline the locked object and calculate the size. The object category, outline and size are all sent to the master control robot, which compares the detected object category, outline and size with the corresponding information stored in the storage unit one by one. If all the information matches successfully, it indicates that the target object verification is successful; if any information does not match, it indicates that the target object verification has failed.

[0026] Furthermore, a weighing machine is provided on the hook for weighing the items lifted by the overhead crane; when the items are put into storage, the weighing machine weighs the items lifted by the overhead crane, and the weighing machine sends the weight to the master control robot, which saves the weight in the attribute information corresponding to the items.

[0027] Furthermore, the method further includes step S5 of searching for a target stacking area when the items are put into storage:

[0028] When items are put into storage, the stacking area for items of the same category is found according to the category of the items. The current height of the stacking area is determined according to the coordinates of the items in the stacking area. The target stacking area is determined according to the current height of the stacking area and the size and weight of the items.

[0029] Furthermore, when placing objects, the master control robot compares the areas of different surfaces according to the size and outline of the objects, and controls the overhead crane to place the side of the object with a larger area flat.

[0030] The present invention identifies the object to be measured and frames the object. The dual millimeter wave radar accurately measures the size of the object by outlining the framed object, so that the laser rangefinder can accurately measure the distance between the outlined object and the laser rangefinder. When the object is put into storage, the category and identification information of the object are recorded, and the coordinates of the object are calculated. When the object is moved, the category of the object can be identified only based on the coordinates, and the coordinates of the object will be updated when the object moves. The attribute information and the object are always corresponding.

[0031] When the overhead crane carries items according to the moving instruction, the master control robot verifies the items carried by the overhead crane and the moving instruction to determine whether the items carried by the overhead crane are correct; it also calculates the size and weight of the items to prevent the items from being swapped during the moving process, identifies the category of the items when the items are about to be moved, calculates the size, weight and coordinates of the items, searches for the stored attribute information according to the coordinates, and determines whether the category, size and weight are the same as the attribute information and the moving instruction obtained by the search. If they are the same as the attribute information, it means that the items have not been swapped, and if they are different, it means that the items have been swapped, and an alarm needs to be issued so that the staff can deal with the problems in time. Compared with the moving instruction, if they are the same, the crane follows until the items are carried to the target position, and if they are different, it means that the system fails, and an alarm is issued or an adjustment instruction is issued. The accuracy of the overhead crane when carrying items is ensured. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a flow chart of the ranging method. DETAILED DESCRIPTION

[0033] The present invention is further described below.

[0034] The full-space precise distance measurement device based on machine vision provided by the present invention comprises a master control robot, a storage unit, a power supply and three distance measurement units, the three distance measurement units are respectively a first distance measurement unit, a second distance measurement unit and a third distance measurement unit, the first distance measurement unit comprises a first chassis, a first laser rangefinder and a first controller, the second distance measurement unit comprises a second chassis, a second laser rangefinder and a second controller, and the third distance measurement unit comprises a third chassis, a camera, a third laser rangefinder, a third controller and a dual millimeter wave radar.

[0035] The chassis is used to carry the components in the corresponding distance measuring unit, and the power supply is used to provide working power for the components of the distance measuring device. The first controller of the first distance measuring unit is connected to the first laser rangefinder, and the first laser rangefinder is set on the panel of the first chassis; the second controller of the second distance measuring unit is connected to the second laser rangefinder, and the second laser rangefinder is set on the panel of the second chassis; the third controller of the third distance measuring unit is respectively connected to the camera, the third laser rangefinder and the dual millimeter wave radar, and the camera, the third laser rangefinder and the dual millimeter wave radar are all set on the panel of the third chassis. At the same time, the first controller, the second controller and the third controller are also connected to the master control robot respectively. The master control robot is used to issue instructions to each controller and receive information sent by each controller, process the information after receiving it, and finally store it in the storage unit. The first controller and the second controller send the distance measurement information to the master control robot. In addition to sending the distance measurement information to the master control robot, the third controller also sends the type, outline, and size of the object to the master control robot. The master control robot can calculate the three-dimensional coordinates of the object based on the distance measurement information sent by the three controllers, and finally store the three-dimensional coordinate information of the object, the type, outline, and size of the object in the storage unit. The camera in the third distance measurement unit is used to obtain the video stream, and the dual millimeter wave radar is used to scan the outline of the object and calculate the size. The third controller contains an object recognition model, which is pre-trained and can be used to identify the category of the object.

[0036] The laser distance meters in the first distance measuring unit and the second distance measuring unit may also be replaced by distance sensors.

[0037] The present invention is described in detail below in combination with the full-space precise distance measurement method based on machine vision, as follows:

[0038] S1. Set up a distance measuring device and install three distance measuring units on the overhead crane.

[0039] For large warehouses, it is often necessary to use an overhead crane to move items. The structure of the overhead crane specifically includes a beam and a trolley arranged on the beam. The beam is arranged on the top of the warehouse, and columns can be arranged at both ends of the beam to support it. Guide rails are arranged at the lower ends of the columns, so that the beam can move with the columns at both ends. Of course, guide rails can also be arranged on the top of the warehouse so that the two ends of the beam can move along the guide rails. At this time, the guide rails on the top of the warehouse need to be supported by corresponding support frames. The trolley is arranged on the beam and can move along the beam. A retractable hook is arranged under the trolley, and the items can be lifted and moved by the hook.

[0040] When setting the distance measuring unit, the first distance measuring unit is set at one end of the beam, the second distance measuring unit is set on the trolley, and the third distance measuring unit is set on the hook under the trolley. The third distance measuring unit moves up and down with the hook. The first laser distance measuring unit in the first distance measuring unit emits laser to illuminate a fixed wall in the warehouse, so as to measure the distance between the beam and the wall. Since the second distance measuring unit is set on the trolley, the second laser distance measuring unit in the second distance measuring unit emits laser to illuminate the iron plate at the end of the beam (the iron plate is preset), so as to measure the distance between the trolley and the end of the beam. The third laser distance measuring unit in the third distance measuring unit is used to measure the distance between the hook and the beam after the item is put down.

[0041] A reference point needs to be set in the warehouse area, and a coordinate system is established with the reference point as the origin.

[0042] When carrying items, first the beam is moved to the set position. After the distance measured by the first laser rangefinder is sent to the master control robot via the first controller, the master control robot can calculate the distance the beam has moved based on the distance, and the horizontal coordinates of the item in the direction of the beam's movement can be obtained. Then the trolley is controlled to move along the beam. After the trolley moves into position, the distance measured by the second laser rangefinder is sent to the master control robot via the second controller. The master control robot can calculate the distance the trolley has moved based on the distance, and the horizontal coordinates of the item in the direction of the trolley's movement can be obtained. After the hook puts the item down, the distance measured by the third laser rangefinder is sent to the master control robot via the third controller. The master control robot can calculate the distance the hook has dropped based on the distance, and the vertical coordinates of the item can be obtained. The three-dimensional coordinate information of the item can then be obtained.

[0043] S2. When the target item enters the storage area, the identification information of the target item is used as the attribute information of the target item and stored in the storage unit, and the master control robot controls the third distance measuring unit to identify the target item and add the identification result to the attribute information of the target item. Among them, the identification information is the name of the owner.

[0044] When the target item enters the warehouse area, the identification information of the target item (i.e., the name of the owner) is added to the attribute information of the target item and stored in the storage unit. The identification information can be transmitted from the robot registering the warehouse to the master control robot, or it can be input by the staff. When the item enters the warehouse area, the third ranging unit needs to identify the target item first, specifically: the video stream of the item is obtained in real time through the camera, and the video stream is transmitted to the third controller, and the item recognition model in the third controller identifies the type of the item. The item recognition model in the third controller is a pre-trained model. A large number of video streams containing items to be identified are trained in the item recognition model in advance. The data set is divided into a training set, a test set, and a validation set. First, the samples in the training set are input into the item recognition model, and the parameters of the model are adjusted. Then, the samples in the validation set are input into the model, and the hyperparameters are adjusted. Finally, the test set is input into the model to evaluate the generalization ability of the model. For example, when the target item is a container, it is necessary to continuously shoot videos of the container and manually annotate the videos. The annotated videos are then input as samples into the object recognition model for training, so that the object recognition model can recognize containers of various sizes.

[0045] The third controller identifies the type of objects in the video stream through the object recognition model therein, and obtains the edge pixels of the target objects in the video stream based on the edge detection technology, connects the edge pixels to form an image annotation frame similar to the size of the object, and frames the object with the image annotation frame to lock the target object.

[0046] The third controller sends the video stream of the locked target object to the dual millimeter-wave radar, and the dual millimeter-wave radar outlines the locked object and measures the size of the target object.

[0047] The dual millimeter-wave radar transmits and receives electromagnetic waves in the millimeter-wave band to outline the locked object and measure its spatial dimensions. The dual millimeter-wave radar can achieve multi-target separation and tracking through multi-target detection and tracking algorithms, thereby outlining the outline of the object, and calculating the size of the target object by calculating the time difference between the transmitted electromagnetic wave and the received reflected electromagnetic wave. The dual millimeter-wave radar transmits the outline and size of the target object to the third controller, which adds the type, outline, and size of the target object to the attribute information through the master control robot and stores it in the storage unit.

[0048] S3. After the overhead crane moves the target object into the storage area, three laser rangefinders measure the distance of the target object and send it to the corresponding controller. The master control robot calculates the three-dimensional coordinates of the target object based on the distances measured by the three laser rangefinders, adds the three-dimensional coordinates of the target object to the attribute information of the target object, and stores it in the storage unit.

[0049] After the target object is transported to the destination by the hook of the trolley, the laser rangefinder in the third distance measuring unit measures its vertical distance from the top of the beam, and the master control robot can calculate the z-axis coordinate of the target object based on the distance. The horizontal coordinates (x-axis and y-axis) of the object can be obtained by the laser rangefinder installed in the first distance measuring unit and the second distance measuring unit. Finally, the master control robot calculates the three-dimensional coordinate information of the object after being transported to the warehouse area by the overhead crane based on the received distance information, and the master control robot stores the three-dimensional coordinate information in the attribute information of the corresponding object in the storage unit.

[0050] Assuming that the target object is a container, the direction parallel to the trolley track is the y direction, the direction perpendicular to the track in the horizontal plane is the x direction, and the vertical direction is the z direction. When the trolley carries the target object to the stacking area of ​​the container and the hook and the object on the trolley are not separated, the third laser rangefinder in the third distance measuring unit measures the vertical distance between it and the beam, and the third laser rangefinder transmits the distance information to the master control robot through the third controller, and the master control robot can calculate the coordinates of the transported container in the z-axis direction; the second laser rangefinder in the second distance measuring unit measures the distance between the trolley and the end of the beam, and the measured distance is uploaded to the master control robot, and the master control robot can calculate the coordinates of the transported container in the y-axis direction according to the distance; the first laser rangefinder in the first distance measuring unit is installed on the beam (can also be installed on the trolley), and is used to measure the distance between the beam and a certain obstacle (such as a certain wall), and the measured distance is uploaded to the master control robot, and the master control robot can calculate the coordinates of the transported container in the x-axis direction based on this; thereby obtaining the three-dimensional coordinates x, y, and z of the container.

[0051] S4. After the target object enters the storage area, when the overhead crane moves the target object according to the moving instruction, the master control robot verifies the target object contacted by the hook, and after the target object moves, recalculates the coordinates of the target object, and updates the coordinates in the attribute information of the target object according to the coordinates after the movement.

[0052] For the target object to be moved, when the hook contacts the target object, the third distance measuring unit measures the vertical distance between the target object and the beam, and sends it to the master control robot through the third controller. The master control robot simultaneously receives the distance measurement information sent by the first controller and the second controller, calculates the current three-dimensional coordinate information of the target object to be moved, and determines whether the current three-dimensional coordinates of the target object are consistent with the coordinates in the moving instruction. If the coordinates are inconsistent, it may be that the overhead crane has found the wrong object to be moved, and adjustment instructions are sent to the overhead crane until the overhead crane finds the object at the correct position.

[0053] For example, the master control robot calculates the coordinates of the hook as (4, 5, 8), and the position of the object in the instruction is (4, 3, 8). Then the overhead crane needs to move 2m in the negative y direction, and the master control robot sends an instruction to the overhead crane to move 2m in the negative y direction.

[0054] When the coordinates of the target object are consistent with the coordinates in the moving instruction, the master control robot controls the third controller to identify and lock the target object, that is, to identify and lock the target object through the object recognition model, and then use the dual millimeter wave radar to outline the locked target object, and calculate the size of the target object, and send the type, outline and size of the target object to the master control robot respectively, and the master control robot compares this information with the item attribute information at the corresponding coordinate point in the storage unit one by one. When all the information is successfully matched, the master control robot issues an instruction to move the target object through the overhead crane. Until the overhead crane moves the target object to the new destination, after the target object is put down, the laser rangefinder of the third distance measuring unit is used again to measure the height of the target object, combined with the distance measured by the first distance measuring unit and the second distance measuring unit in the horizontal direction, and finally the master control counter calculates the new coordinates of the target object after moving, and updates the new coordinates of the target object after moving in the attribute information of the target object.

[0055] Therefore, the coordinate information of the items moving in the storage area will be updated in time, and the coordinate information will be stored in the storage unit together with its identification information. Therefore, even for items with the same appearance (such as containers of the same size and shape), even if the items are moved multiple times in the storage area, they can still be accurately identified by the device of the present invention without confusion.

[0056] When moving items in the warehouse area, the three distance measuring units first measure the coordinates of the items to be moved and compare them with the coordinate information in the moving instruction to prevent the crane from moving incorrectly. After the position is correct, the third controller will identify the type of the item and outline it, and compare it with the corresponding attribute information in the storage unit to prevent the item from being swapped. Because for the items in the warehouse area, the appearance of the items is similar in many cases, and it is impossible to accurately judge by the human eye. The present invention first performs type identification through the third controller and then uses the dual millimeter wave radar to identify the outline and size of the item. When all the information is matched successfully, it means that the item has not been moved and has not been swapped. If the three-dimensional coordinate information matches, the item type matches, but the item outline and size do not match, it means that the item is a swapped item. If the item type, item outline, and size match, but the three-dimensional coordinate information does not match, it means that the item has been moved. If the three-dimensional coordinate information does not match, and the item type, item outline, and size do not match, it means that the crane has moved incorrectly and needs to reach the specified position again. The present invention can also be provided with an alarm, and once it is found that the information matching fails, the alarm is issued by the alarm.

[0057] The present invention can identify containers and carry containers. The moving instructions are: a cargo owner, container, 1.2m×1.2m×1.2m, coordinates (2, 4.5, 6), outline a, target position (12, 23, 5). The overhead crane moves to the top of the item, the hook contacts the item, and the master control robot calculates the coordinates of the item according to the distances measured by the first distance measuring unit, the second distance measuring unit and the third distance measuring unit; it is determined whether the calculated coordinates are the same as the coordinates (2, 4.5, 6) in the moving instruction. If they are the same, the third distance measuring unit identifies the item to obtain the category of the item, outlines the item, and obtains the size of the item. The above measured information is compared with the information in the moving instruction to determine whether they are the same (this process can be determined by the third controller or the master control robot). If they are the same, the overhead crane performs the carrying task and carries the container to the target position (12, 23, 5). The distance measuring device tracks the overhead crane to the target position, measures the coordinates of the container after movement, and stores them.

[0058] When items enter the storage area, they can also be weighed and the weight stored in the attribute information.

[0059] At the same time, in order to prevent items in the warehouse from being moved by people, several third ranging units are set on the walls of the warehouse to monitor the status of the items and whether the items will be moved without instructions. If an abnormal state occurs, an alarm message will be issued.

[0060] The working process of the present invention is further described in detail below.

[0061] When items are put into storage, the master control robot can find the stacking area for items of the same category according to the category of the items, determine the current height of the stacking area according to the coordinates of the items in the stacking area, and determine the target stacking area according to the current height of the stacking area and the size and weight of the items.

[0062] When moving items, you can determine in advance where the items will be placed. By judging the height of the items currently placed in each stacking area and the number of items in this batch, and comparing them with the height of the warehouse, you can determine whether the batch of items can be placed. Try to place each batch in one stacking area.

[0063] The master control robot determines the stacking area of ​​the same category as the item, determines the available stacking area among the stacking areas of the same category according to the size and weight of the items already placed in the stacking area, determines the height of the stacking area according to the location of the stacking area and the coordinates of the items in the stacking area, further determines the stacking area to be placed in the batch of items according to the number of items in the batch and the size of the items, and sends the location of the target stacking area to the overhead crane.

[0064] When picking up goods in the warehouse, generally the same category of items are picked up. Therefore, when placing the items, it is preferred to place the same category of items in the same stacking area. In order to avoid placing overweight items on top of lighter items and squeezing the lighter items, it is necessary to select a stacking area where the difference between the weight of the placed items and the weight of the items to be placed is within the preset difference range. In order to further ensure the stability of the stacking area, it is necessary to select a stacking area where the difference between the size of the placed items and the size of the items to be placed is within the preset size difference range. When placing each batch of items, try to place the same batch of items in the same stacking area. For example, there are 200 items in batch A to be stored. By calculation, it can be known that all the items of batch A can be placed in the B1 stacking area, B2 can place 80 items of batch A, and B3 can place 150 items of batch A. Then batch A will be placed in the B1 stacking area.

[0065] When placing items, you can arrange them through reasonable space planning to make more space for the items. For example, laying the larger side of the item flat can increase the stability of the stacking. The master control robot calculates the area of ​​each side based on the size and outline of the item, and compares the area of ​​each side to determine the side on which the item should be laid flat when it is put down.

[0066] When stacking goods in traditional warehouses, a certain amount of space needs to be reserved above the stacking area for workers to check. The present invention only needs to use a distance measuring device to check the placement of the goods and command the overhead crane to place the items. There is no need to reserve a large space above the stacking area, thus maximizing the use of the warehouse space.

Claims

1. A full-space precise distance measurement device based on machine vision, characterized in that: include: The first distance measuring unit is connected to the master control robot and the power supply, respectively, and is installed on the crossbeam of the overhead travelling crane, and includes a chassis, a first laser distance meter and a first controller, and is used to measure the horizontal distance between the crossbeam and a specific wall surface; The second distance measuring unit is connected to the master control robot and the power supply respectively, and is installed on the mobile trolley of the overhead travelling crane, and includes a chassis, a second laser distance meter, and a second controller, and is used to measure the horizontal distance between the mobile trolley and the shielding object at the end of the beam; The third distance measuring unit is connected to the master control robot and the power supply respectively, and is installed on the hook of the mobile vehicle, and includes a chassis, a camera, a third laser rangefinder, a third controller and a dual millimeter wave radar, and is used to measure the vertical distance between the hook and the beam; The master control robot is connected to the storage unit, the power supply and the three distance measuring units respectively, and is used to control the three controllers to work, receive the distance measuring information sent by each controller, and measure the three-dimensional coordinate information of the target object according to the received distance measuring information; A storage unit connected to the master control robot and used to store attribute information of the item; and The power supply is respectively connected to the master control robot, the storage unit and the three distance measuring units to provide working power for each part.

2. The full-space precise distance measurement device based on machine vision according to claim 1, characterized in that: The first laser rangefinder and the second laser rangefinder are respectively replaced by distance sensors.

3. A full-space precise distance measurement method based on machine vision, characterized in that: The steps include: S1. Set up the distance measuring device as claimed in claim 1 or 2, wherein the first distance measuring unit is set on the beam to measure the distance between the beam and the wall, the second distance measuring unit is set on the trolley to measure the distance between the trolley and the end of the beam, and the third distance measuring unit is set on the hook to measure the height distance between the hook and the beam in the vertical direction when the hook contacts the object; set a reference point in the warehouse area, and establish a coordinate system with the reference point as the origin; S2. When the target item enters the storage area, the identification information of the target item is used as the attribute information of the target item and stored in the storage unit. The master control robot controls the third ranging unit to identify the target item and adds the identification result to the attribute information of the target item, wherein the identification information is the name of the owner; S3. After the overhead crane carries the target object to the storage area, the three laser rangefinders measure the distance and send it to the master control robot via the corresponding controller. The master control robot calculates the coordinates of the target object based on the distance measured by the three laser rangefinders, adds the coordinates of the target object to the attribute information of the target object, and stores it in the storage unit; S4. After the target object enters the storage area, when the overhead crane moves the target object according to the moving instruction, the master control robot verifies the target object contacted by the hook, and after the target object moves, recalculates the coordinates of the target object, and updates the coordinates in the attribute information of the target object according to the coordinates after the movement.

4. The method for accurate distance measurement in full space based on machine vision according to claim 3, characterized in that: In step S2, the master control robot controls the third distance measuring unit to identify the target object in the following specific manner: S2a-1. The master control robot sends an object recognition instruction to the third controller, and the third controller controls the camera to obtain a video stream of the object, and inputs the video stream obtained by the camera into the object recognition model therein, identifies and locks the target object, and obtains the category of the object; S2a-2. The third controller sends the video stream of the locked target object to the dual millimeter wave radar, and the dual millimeter wave radar outlines the locked object and measures the size of the object.

5. The method for accurate distance measurement in full space based on machine vision according to claim 3, characterized in that: The specific method of verifying the target object contacted by the hook in step S4 is: S4a-1. When the hook contacts the object, the three laser rangefinders measure the distance and send it to the master control robot through their respective corresponding controllers. The master control robot calculates the three-dimensional coordinates of the object contacted by the hook based on the distance measurement information; if the calculated three-dimensional coordinates are different from the coordinates in the movement instruction, the master control robot sends an adjustment instruction to the overhead traveling crane until the coordinates of the object contacted by the hook are the same as the coordinates in the movement instruction; if the coordinates of the object contacted by the hook are the same as the coordinates in the movement instruction, execute step S4a-2; S4a-2. The third controller controls the camera to obtain the video stream of the object, and inputs the video stream to the object recognition model. The object recognition model identifies the type of object, and the third controller locks the identified object. Then the third controller controls the dual millimeter-wave radar to outline the locked object and calculate the size. The object category, outline and size are all sent to the master control robot, which compares the detected object category, outline and size with the corresponding information stored in the storage unit one by one. If all the information matches successfully, it indicates that the target object verification is successful; if any information does not match, it indicates that the target object verification has failed.

6. The method for accurate distance measurement in full space based on machine vision according to claim 3, characterized in that: A weighing machine is also provided on the hook to weigh the items lifted by the overhead crane. When the items are put into storage, the weighing machine weighs the items lifted by the overhead crane, and the weighing machine sends the weight to the master control robot, which saves the weight in the attribute information corresponding to the items.

7. The method for accurate distance measurement in full space based on machine vision according to claim 6, characterized in that: The step S5 is also included to find the target stacking area when the items are put into storage: When items are put into storage, the stacking area for items of the same category is found according to the category of the items. The current height of the stacking area is determined according to the coordinates of the items in the stacking area. The target stacking area is determined according to the current height of the stacking area and the size and weight of the items.

8. The method for accurate distance measurement in full space based on machine vision according to claim 3, characterized in that: When placing items, the master control robot compares the areas of different surfaces according to the size and outline of the items, and controls the overhead crane to place the larger side of the item flat.

Images