Energy-saving handling method for building construction materials, computer equipment and storage medium

Abstract

Embodiments of the present disclosure disclose an energy-saving handling method for building construction materials, a computer device, and a storage medium. A specific implementation manner of this method includes: generating an initial loss value and a prediction loss value based on a reset image sample data set; generating a sample loss value according to the initial loss value and the above prediction loss value; generating a target image sample data set according to the initial image sample data set and a sample weight set; training an initial building construction material pixel segmentation model according to the target image sample data set to obtain a building construction material pixel segmentation model; generating at least one set of planar grasping information and grasping order information corresponding to the building construction materials according to a cargo box image and the building construction material pixel segmentation model; and controlling a material handling device to transport each building construction material in a building construction material cargo box to a preset position according to the set of planar grasping information and the grasping order information. This implementation manner can detect abnormal conditions of the transformer in a timely manner.

Description

Technical Field

[0001] Embodiments of the present disclosure relate to the field of building material handling, and particularly to an energy-saving handling method for building construction materials, computer equipment, and a storage medium. Background Art

[0002] Currently, at a construction site, a large number of building materials need to be transported. At present, in order to transport building materials more quickly, the commonly adopted method is as follows: through an image recognition model, identify the positions of building materials, and then control a handling device to handle each building material. However, when adopting the above method, the following technical problems usually exist: 1. A large amount of image style data is required to train the image recognition model, resulting in a relatively large operating pressure on the computing device; 2. Only determining the positions of building materials through images is likely to cause inaccurate position recognition, resulting in ineffective handling by the handling device and wasting handling resources. Summary of the Invention

[0003] This part of the content of the present disclosure is used to briefly introduce concepts, which will be described in detail in the subsequent Detailed Description part. This part of the content of the present disclosure is not intended to identify the key features or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.

[0004] Some embodiments of the present disclosure propose an energy-saving handling method for building construction materials, computer equipment, and a computer-readable storage medium to solve the technical problems mentioned in the above Background Art part.

[0005] In a first aspect, some embodiments of the present disclosure provide an energy-saving handling method for building construction materials. The method includes: obtaining an initial image sample data set and a sample weight set; generating a reset image sample data set according to the initial image sample data set and the sample weight set; generating an initial loss value and a predicted loss value based on the reset image sample data set; generating a sample loss value according to the initial loss value and the predicted loss value; if the sample loss value is less than or equal to a preset loss value, generating a target image sample data set according to the initial image sample data set and the sample weight set; training an initial building construction material pixel segmentation model according to the target image sample data set to obtain a building construction material pixel segmentation model; collecting a box image of a current building construction material box to be handled, where the box image shows at least one building construction material; generating a plane grasping information group and a grasping order information corresponding to the at least one building construction material according to the box image and the building construction material pixel segmentation model; controlling an associated material handling device to handle each building construction material in the building construction material box to a preset position according to the plane grasping information group and the grasping order information.

[0006] Optionally, generating the planar grasping information group and the grasping order information corresponding to the at least one building construction material according to the cargo box image and the building construction material pixel segmentation model described above includes: inputting the cargo box image into the building construction material pixel segmentation model to obtain a pixel segmentation result group; using the pixel segmentation result group to generate point cloud information corresponding to each building construction material to obtain a point cloud information group; using the point cloud information group to generate planar grasping information corresponding to each building construction material to obtain a planar grasping information group; and generating the grasping order information of each building construction material by using the position information of each building construction material in the cargo box image.

[0007] Optionally, generating the reset image sample data set according to the initial image sample data set and the sample weight set described above includes: generating an image sample data vector of each initial image sample data in the initial image sample data set to obtain an image sample data vector set; and using the sample weight set and the image sample data vector set to generate the reset image sample data set.

[0008] Optionally, training the initial building construction material pixel segmentation model according to the target image sample data set to obtain the building construction material pixel segmentation model includes: based on the target image sample data set, performing the following processing steps: inputting the target image sample data set into the initial building construction material pixel segmentation model to obtain a sample image segmentation result set; generating a sample error value based on the sample image segmentation result set and each sample label corresponding to the target image sample data set; in response to determining that the sample error value is less than or equal to a preset threshold, determining the initial model as the trained building construction material pixel segmentation model; and in response to determining that the sample error value is greater than the preset threshold, adjusting the network parameters of the initial building construction material pixel segmentation model according to the sample error value and the preset threshold to perform the above processing steps again.

[0009] Optionally, using the point cloud information group to generate planar grasping information corresponding to each building construction material includes: inputting the point cloud information corresponding to the building construction material into a pre-trained planar detection neural network model to obtain at least one initial planar grasping information; and determining the initial planar grasping information that meets the preset grasping conditions in the at least one initial planar grasping information as the planar grasping information.

[0010] Optionally, generating the grasping sequence information of the above-mentioned construction materials by using the position information of each construction material in the above-mentioned cargo box image includes: determining the height information of each construction material according to the position information of each construction material; sorting each construction material from high to low according to the height represented by each height information, and in the order from the exit to the interior of the above-mentioned construction material cargo box, to obtain a construction material sequence as the grasping sequence information.

[0011] In a second aspect, the present disclosure also provides a computer device, where the computer device includes a processor, a memory, and a computer program stored on the memory and executable by the processor. When the computer program is executed by the processor, the method described in any implementation manner of the first aspect is implemented.

[0012] In a third aspect, the present disclosure also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the method described in any implementation manner of the first aspect is implemented.

[0013] The above - mentioned various embodiments of the present disclosure have the following beneficial effects: Through the energy - saving handling method of building construction materials in some embodiments of the present disclosure, the operating pressure of the computing device is reduced, and the efficiency of handling building construction materials is improved. First, an initial image sample data set and a sample weight set are obtained. Thus, it is convenient to train the building construction material pixel segmentation model. Secondly, according to the above - mentioned initial image sample data set and the above - mentioned sample weight set, a reset image sample data set is generated. Thus, it is convenient to update the sample weight set subsequently. Thirdly, based on the above - mentioned reset image sample data set, an initial loss value and a prediction loss value are generated. Thus, the error between the predicted output of the initial image sample data set and the sample weight set can be determined. Therefore, not only the speed of the subsequent model learning the target image sample data set is improved, but also the accuracy of the model is improved. Then, according to the above - mentioned initial loss value and the above - mentioned prediction loss value, a sample loss value is generated. Then, if the above - mentioned sample loss value is less than or equal to a preset loss value, according to the above - mentioned initial image sample data set and the sample weight set, a target image sample data set is generated. Thus, the data volume of the sample data can be controlled by the preset loss value, shortening the training duration of the model to improve the training speed of the model. Then, according to the above - mentioned target image sample data set, the initial building construction material pixel segmentation model is trained to obtain the building construction material pixel segmentation model. After that, a box image of the building construction material box to be currently handled is collected. Among them, the above - mentioned box image shows at least one building construction material. Then, according to the above - mentioned box image and the above - mentioned building construction material pixel segmentation model, a plane grasping information group and a grasping order information corresponding to the above - mentioned at least one building construction material are generated. Finally, according to the above - mentioned plane grasping information group and the above - mentioned grasping order information, the associated material handling equipment is controlled to move each building construction material in the above - mentioned building construction material box to a preset position. Thus, the operating pressure of the computing device is reduced, and the efficiency of handling building construction materials is improved. Description of the Drawings

[0014] In combination with the accompanying drawings and with reference to the following specific embodiments, the above - mentioned and other features, advantages, and aspects of the various embodiments of the present disclosure will become more obvious. Throughout the drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic, and the elements and elements are not necessarily drawn to scale.

[0015] Figure 1 is a flowchart of some embodiments of the energy - saving handling method of building construction materials according to the present disclosure;

[0016] Figure 2 is a schematic block diagram of the structure of a computer device provided by an embodiment of the present disclosure. Detailed Embodiments

[0017] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although some embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. On the contrary, these embodiments are provided to more thoroughly and completely understand the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not used to limit the protection scope of the present disclosure.

[0018] In addition, it should be noted that for the sake of convenience of description, only parts related to the relevant invention are shown in the drawings. Without conflict, the embodiments in the present disclosure and the features in the embodiments can be combined with each other.

[0019] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence relationship of the functions performed by these devices, modules or units.

[0020] It should be noted that the modifications of "one" and "multiple" mentioned in the present disclosure are illustrative rather than restrictive. Those skilled in the art should understand that unless otherwise clearly specified in the context, it should be understood as "one or more".

[0021] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only for illustrative purposes and are not used to limit the scope of these messages or information.

[0022] The present disclosure will be described in detail below with reference to the drawings and in combination with embodiments.

[0023] Figure 1 is a flowchart of some embodiments of an energy-saving handling method for building construction materials according to the present disclosure. The process 100 of some embodiments of the energy-saving handling method for building construction materials according to the present disclosure is shown. The energy-saving handling method for building construction materials includes the following steps:

[0024] Step 101, obtaining an initial image sample data set and a sample weight set.

[0025] In some embodiments, the execution entity (such as a computing device) of the energy-saving handling of building construction materials can obtain the initial image sample data set and the sample weight set from the terminal by means of a wired connection or a wireless connection.

[0026] Among them, the initial image sample data in the above initial image sample dataset can be data for training the initial building construction material pixel segmentation model. For example, the initial image sample data in the above initial image sample dataset can be a cargo box image. The sample weights in the above sample weight set can characterize the importance of the initial image sample data for model training. The larger the sample weight, the greater the impact of the initial image sample data on model training. One initial image sample data corresponds to one sample weight. The initial building construction material pixel segmentation model can be an untrained residual neural network model.

[0027] Step 102: Generate a reset image sample dataset according to the above initial image sample dataset and the above sample weight set.

[0028] In some embodiments, the above execution subject can generate a reset image sample dataset according to the above initial image sample dataset and the above sample weight set.

[0029] In an actual application scenario, the above execution subject can generate a reset image sample dataset through the following steps:

[0030] First step: Generate an image sample data vector for each initial image sample data in the above initial image sample dataset to obtain an image sample data vector set. For example, each initial image sample data can be input into a vector encoding neural network to obtain an image sample data vector. Such as, the vector encoding neural network can be a Bert network or a convolutional neural network.

[0031] Second step: Use the above sample weight set and the above image sample data vector set to generate a reset image sample dataset. For example, the product of each sample weight and the image sample data vector corresponding to the above sample weight can be determined as the reset image sample data to obtain a reset image sample dataset.

[0032] Step 103: Generate an initial loss value and a prediction loss value based on the above reset image sample dataset.

[0033] In some embodiments, the above-mentioned execution entity may generate an initial loss value and a prediction loss value based on the above-mentioned reset image sample dataset. The initial loss value may represent the inner product value between the gradients of the function corresponding to the initial building construction material pixel segmentation model with respect to the model parameters. The prediction loss value may represent the difference between the output value of the initial building construction material pixel segmentation model and the true value. For example, the above-mentioned execution entity may first input the above-mentioned reset image sample dataset into the initial building construction material pixel segmentation model to obtain a model prediction value corresponding to the above-mentioned reset image sample dataset. Then, the above-mentioned execution entity may determine the loss value between the above-mentioned model prediction value and the true value corresponding to the reset image sample dataset. Finally, using the neural network kernel function, the initial loss value corresponding to the above-mentioned reset image sample dataset is determined.

[0034] Step 104: Generate a sample loss value according to the above-mentioned initial loss value and the above-mentioned prediction loss value.

[0035] In some embodiments, the above-mentioned execution entity may generate a sample loss value according to the above-mentioned initial loss value and the above-mentioned prediction loss value. The average value of the above-mentioned initial loss value and the above-mentioned prediction loss value may be determined as the sample loss value.

[0036] Step 105: If the above-mentioned sample loss value is less than or equal to a preset loss value, generate a target image sample dataset according to the above-mentioned initial image sample dataset and the sample weight set.

[0037] In some embodiments, if the above-mentioned sample loss value is less than or equal to a preset loss value, the above-mentioned execution entity may generate a target image sample dataset according to the above-mentioned initial image sample dataset and the sample weight set. First, the sample weights in the above-mentioned sample weight set that are greater than or equal to the preset weight may be determined as target sample weights to obtain a target sample weight set. Then, each initial image sample data corresponding to the above-mentioned target sample weight set may be determined as the target image sample dataset. There is no limitation on the setting of the preset weight.

[0038] Step 106: Train the initial building construction material pixel segmentation model according to the above-mentioned target image sample dataset to obtain a building construction material pixel segmentation model.

[0039] In some embodiments, the above-mentioned execution entity may train the initial building construction material pixel segmentation model according to the above-mentioned target image sample dataset to obtain a building construction material pixel segmentation model.

[0040] In an actual application scenario, the above-mentioned execution entity may train the initial building construction material pixel segmentation model through the following steps to obtain a building construction material pixel segmentation model:

[0041] In the first step, based on the above-mentioned target image sample dataset, perform the following processing steps:

[0042] 1. Input the target image sample dataset into the above-mentioned initial building construction material pixel segmentation model to obtain a sample image segmentation result set. The above-mentioned sample image segmentation result set can be the result set output by the initial building construction material pixel segmentation model. The sample image segmentation result can refer to the result of pixel-level segmentation of each building material included in the target image sample data.

[0043] 2. Generate a sample error value based on the above-mentioned sample image segmentation result set and each sample label corresponding to the above-mentioned target image sample dataset. First, it is possible to determine the number of labels for which the sample image segmentation result in the above-mentioned sample image segmentation result set is different from the corresponding sample label. Then, the ratio of the number of labels to the number of sample image segmentation results included in the sample image segmentation result set is used as the sample error value.

[0044] 3. In response to determining that the above-mentioned sample error value is less than or equal to a preset threshold, determine the above-mentioned initial model as the trained building construction material pixel segmentation model.

[0045] In the second step, in response to determining that the above-mentioned sample error value is greater than the above-mentioned preset threshold, adjust the network parameters of the above-mentioned initial building construction material pixel segmentation model according to the above-mentioned sample error value and the above-mentioned preset threshold to perform the above-mentioned processing steps again.

[0046] Step 107, collect the box image of the building construction material box to be transported currently.

[0047] In some embodiments, the above-mentioned execution entity can collect the box image of the building construction material box to be transported currently. For example, the box image of the building construction material box to be transported currently can be collected through a camera. The building construction material box stores at least one building construction material. For example, the building construction material can be materials such as tiles and cement.

[0048] Step 108, generate the plane grasping information group and grasping sequence information corresponding to the above-mentioned at least one building construction material according to the above-mentioned box image and the above-mentioned building construction material pixel segmentation model.

[0049] In some embodiments, the above-mentioned execution entity can generate the plane grasping information group and grasping sequence information corresponding to the above-mentioned at least one building construction material according to the above-mentioned box image and the above-mentioned building construction material pixel segmentation model.

[0050] In an actual application scenario, the above-mentioned execution entity can generate the plane grasping information group and grasping sequence information corresponding to the above-mentioned at least one building construction material through the following steps:

[0051] First step, input the above cargo box image into the above building construction material pixel segmentation model to obtain a pixel segmentation result group. The building construction material pixel segmentation model can be a model that divides a cargo box image into individual building construction materials with pixels as the segmentation unit. For example, the building construction material pixel segmentation model can be a trained residual neural network model. The pixel segmentation result can represent the pose information of the building construction materials in the cargo box image.

[0052] Second step, use the above pixel segmentation result group to generate point cloud information corresponding to each building construction material, obtaining a point cloud information group. For example, the pixel segmentation result corresponding to the above building construction materials can be determined. The position information of the above building construction materials in the building construction material cargo box can be determined using the pixel segmentation result. Subsequently, a point cloud image corresponding to the above position information can be collected by a point cloud camera as the point cloud information.

[0053] Third step, use the above point cloud information group to generate planar grasping information corresponding to each building construction material, obtaining a planar grasping information group.

[0054] The above third step can include:

[0055] 1. Input the point cloud information corresponding to the above building construction materials into a pre-trained planar detection neural network model to obtain at least one initial planar grasping information. The planar detection neural network model can be a point cloud planar detection model based on RSANC.

[0056] 2. Determine the initial planar grasping information that meets the preset grasping conditions among the above at least one initial planar grasping information as the planar grasping information. The preset grasping condition can be: the planar flatness represented by the initial planar grasping information is the largest.

[0057] Fourth step, use the position information of each building construction material in the above cargo box image to generate the grasping sequence information of each of the above building construction materials.

[0058] The above fourth step can include the following steps:

[0059] 1. Determine the height information of each of the above building construction materials according to the position information of each of the above building construction materials. The height information can represent the current height of the building construction material.

[0060] 2. Sort each of the above building construction materials in descending order according to the height represented by each height information and in the order from the exit to the inside of the above building construction material cargo box to obtain a building construction material sequence as the grasping sequence information.

[0061] Regarding the statement in the background art that "merely determining the position of building materials through images is likely to result in inaccurate position recognition, leading to ineffective handling by the handling device and wasting handling resources.", it can be solved through the following steps: First, input the above-mentioned cargo box image into the above-mentioned building construction material pixel segmentation model to obtain a group of pixel segmentation results. Thus, by determining the pixel segmentation results of each building construction material in the cargo box image, the pose information of each building construction material in the building construction material cargo box can be obtained. Second, use the above-mentioned group of pixel segmentation results to generate point cloud information corresponding to each building construction material, obtaining a group of point cloud information. Thus, it is convenient for subsequent handling based on the three-dimensional features represented by the point cloud data. Then, use the above-mentioned group of point cloud information to generate plane grasping information corresponding to each building construction material, obtaining a group of plane grasping information. Thus, through the plane grasping information, it is convenient for subsequent material handling equipment to control the robotic arm to grasp materials. Finally, use the position information of each building construction material in the above-mentioned cargo box image to generate the grasping sequence information of each building construction material. Thus, each building construction material can be grasped accurately and efficiently. Thereby, ineffective handling by the handling device is avoided, and waste of handling resources is avoided.

[0062] Step 109: According to the above-mentioned group of plane grasping information and the above-mentioned grasping sequence information, control the associated material handling equipment to move each building construction material in the above-mentioned building construction material cargo box to a preset position.

[0063] In some embodiments, the above-mentioned execution subject can control the associated material handling equipment to move each building construction material in the above-mentioned building construction material cargo box to a preset position according to the above-mentioned group of plane grasping information and the above-mentioned grasping sequence information. For example, it can control the handling robot connected by communication to move each building construction material in the above-mentioned building construction material cargo box to a preset position according to the grasping sequence information. Here, the preset position can refer to the position preset for placing building materials. The material handling equipment can refer to an intelligent logistics vehicle with a robotic arm.

[0064] Figure 2 The structural schematic block diagram of a computer device provided by an embodiment of the present disclosure. This computer device can be a terminal.

[0065] As Figure 2 shown, this computer device includes a processor, a memory, and a network interface connected through a system bus. Among them, the memory can include a non-volatile storage medium and an internal memory.

[0066] The non-volatile storage medium can store an operating system and a computer program. This computer program includes program instructions, which when executed, can cause the processor to execute any one of the building construction material energy-saving handling methods.

[0067] The processor is used to provide computing and control capabilities to support the operation of the entire computer device.

[0068] The internal memory provides an environment for the operation of a computer program in a non-volatile storage medium. When the computer program is executed by the processor, the processor can execute any one of the energy-saving handling methods for building construction materials.

[0069] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art can understand that Figure 2 the structure shown in is only a block diagram of some structures related to the solution of the present disclosure, and does not constitute a limitation on the computer device to which the solution of the present disclosure is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0070] It should be understood that the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0071] Among them, in one embodiment, the above-mentioned processor is used to run a computer program stored in a memory to implement the following steps: obtaining an initial image sample data set and a sample weight set; generating a reset image sample data set according to the above-mentioned initial image sample data set and the above-mentioned sample weight set; generating an initial loss value and a prediction loss value based on the above-mentioned reset image sample data set; generating a sample loss value according to the above-mentioned initial loss value and the above-mentioned prediction loss value; if the above-mentioned sample loss value is less than or equal to a preset loss value, generating a target image sample data set according to the above-mentioned initial image sample data set and the sample weight set; training an initial building construction material pixel segmentation model according to the above-mentioned target image sample data set to obtain a building construction material pixel segmentation model; collecting a box image of a current building construction material box to be carried, where the above-mentioned box image shows at least one building construction material; generating a plane grasping information group and a grasping order information corresponding to the above-mentioned at least one building construction material according to the above-mentioned box image and the above-mentioned building construction material pixel segmentation model; controlling an associated material handling device to carry each building construction material in the above-mentioned building construction material box to a preset position according to the above-mentioned plane grasping information group and the above-mentioned grasping order information.

[0072] An embodiment of the present disclosure further provides a computer-readable storage medium, on which a computer program is stored. The computer program includes program instructions, and the method implemented when the program instructions are executed can refer to various embodiments of the energy-saving handling method for building construction materials of the present disclosure.

[0073] Among them, the above-mentioned computer-readable storage medium may be an internal storage unit of the above-mentioned computer device in the foregoing embodiment, such as a hard disk or memory of the above-mentioned computer device. The above-mentioned computer-readable storage medium may also be an external storage device of the above-mentioned computer device, such as a plug-in hard disk equipped on the above-mentioned computer device, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc.

[0074] It should be noted that in this article, the terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. Without further limitations, an element defined by the statement "including a..." does not exclude the existence of another identical element in the process, method, article or system including the element.

[0075] The serial numbers of the above embodiments of the present disclosure are only for description and do not represent the superiority or inferiority of the embodiments. As described above, it is only the specific implementation manner of the present disclosure, but the protection scope of the present disclosure is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present disclosure can easily think of various equivalent modifications or substitutions, and these modifications or substitutions should all be covered within the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure shall be subject to the protection scope of the claims.

Claims

1. An energy-saving handling method for building construction materials, characterized in that, The method includes: Obtaining an initial image sample data set and a sample weight set, where the sample weights in the sample weight set represent the importance of the initial image sample data for model training, and one initial image sample data corresponds to one sample weight; Generating a reset image sample data set according to the initial image sample data set and the sample weight set; Generating an initial loss value and a prediction loss value based on the reset image sample data set, where the initial loss value represents the inner product value between the gradients of the function corresponding to the initial building construction material pixel segmentation model with respect to the model parameters; Generating a sample loss value according to the initial loss value and the prediction loss value; If the sample loss value is less than or equal to a preset loss value, generating a target image sample data set according to the initial image sample data set and the sample weight set; Training an initial building construction material pixel segmentation model according to the target image sample data set to obtain a building construction material pixel segmentation model; Collecting a bin image of the current building construction material bin to be carried, where the bin image shows at least one building construction material; Generating a planar grasping information group and a grasping order information corresponding to the at least one building construction material according to the bin image and the building construction material pixel segmentation model; Controlling an associated material handling device to carry each building construction material in the building construction material bin to a preset position according to the planar grasping information group and the grasping order information; Among them, the generating a reset image sample data set according to the initial image sample data set and the sample weight set includes: Generating an image sample data vector for each initial image sample data in the initial image sample data set to obtain an image sample data vector set; Generating a reset image sample data set by using the sample weight set and the image sample data vector set.

2. The method according to claim 1, characterized in that, The generating a planar grasping information group and a grasping order information corresponding to the at least one building construction material according to the bin image and the building construction material pixel segmentation model includes: Inputting the bin image into the building construction material pixel segmentation model to obtain a pixel segmentation result group; Generating point cloud information corresponding to each building construction material by using the pixel segmentation result group to obtain a point cloud information group; Generating planar grasping information corresponding to each building construction material by using the point cloud information group to obtain a planar grasping information group; Generating the grasping order information of each building construction material by using the position information of each building construction material in the bin image.

3. The method according to claim 1, characterized in that, The training an initial building construction material pixel segmentation model according to the target image sample data set to obtain a building construction material pixel segmentation model includes: Based on the target image sample data set, performing the following processing steps: Inputting the target image sample data set into the initial building construction material pixel segmentation model to obtain a sample image segmentation result set; Generating a sample error value based on the sample image segmentation result set and each sample label corresponding to the target image sample data set; In response to determining that the sample error value is less than or equal to a preset threshold, determining the initial building construction material pixel segmentation model as a trained building construction material pixel segmentation model; In response to determining that the sample error value is greater than the preset threshold, adjusting the network parameters of the initial building construction material pixel segmentation model according to the sample error value and the preset threshold to perform the processing step again.

4. The method according to claim 2, characterized in that, The generating of the planar grasping information corresponding to each building construction material by using the point cloud information group includes: Inputting the point cloud information corresponding to the building construction material into a pre-trained planar detection neural network model to obtain at least one initial planar grasping information; Determining the initial planar grasping information that meets the preset grasping conditions among the at least one initial planar grasping information as the planar grasping information.

5. The method according to claim 2, characterized in that, The generating of the grasping sequence information of each building construction material by using the position information of each building construction material in the cargo box image includes: Determining the height information of each building construction material according to the position information of each building construction material; Sorting each building construction material in descending order of the height represented by each height information and in the order from the exit to the inside of the building construction material cargo box to obtain a building construction material sequence as the grasping sequence information.

6. A computer device, wherein, The computer device includes a processor, a memory, and a computer program stored on the memory and executable by the processor, wherein when the computer program is executed by the processor, the steps of the method according to any one of claims 1-5 are implemented.

7. A computer-readable storage medium, wherein, A computer program is stored on the computer-readable storage medium, wherein when the computer program is executed by a processor, the steps of the method according to any one of claims 1-5 are implemented.

Images