阅读说明：本技术 堆积物整体体积的检测方法、装置及电子设备 (Method and device for detecting overall volume of deposit and electronic equipment ) 是由 汤斐 于 2021-02-04 设计创作，主要内容包括：本申请涉及堆积物整体体积的检测方法、装置及电子设备,属于物流技术领域,本申请的方法包括,预先获取目标检测区域的第一图像并对其进行标注处理,确定该目标检测区域的额定立方体边框、并配置该额定立方体的额定方数；基于获取所述第一图像时的拍摄机位及参数、实时获取所述目标检测区域的第二图像,对所述第二图像进行堆积元素检测并基于检测结果进行堆积物的体积立方体绘制,得到实际立方体边框；根据实际立方体边框和额定立方体边框的像素尺寸数据,计算体积立方体与额定立方体的体积比值,将体积比值与额定方数的乘积确定为第二图像中堆积物的体积检测值。本申请有助于更好的实现物流分拣场地中堆积快递的整体体积检测。(The method comprises the steps of obtaining a first image of a target detection area in advance, carrying out labeling processing on the first image, determining a rated cube frame of the target detection area, and configuring the rated square number of the rated cube; acquiring a second image of the target detection area in real time based on the camera position and parameters when the first image is acquired, detecting stacking elements of the second image, and drawing a volume cube of a stacking object based on a detection result to obtain an actual cube frame; and calculating the volume ratio of the volume cube to the rated cube according to the pixel size data of the actual cube border and the rated cube border, and determining the product of the volume ratio and the rated square number as the volume detection value of the accumulated object in the second image. This application helps piling up the whole volume detection of express delivery in better realization commodity circulation letter sorting place.)

1. A method for detecting the entire volume of a heap, comprising:

the method comprises the steps of obtaining a first image of a target detection area in advance, carrying out labeling processing on the first image, determining a rated cube frame of the target detection area, and configuring the rated square number of the rated cube;

acquiring a second image of the target detection area in real time based on the camera position and parameters when the first image is acquired, detecting stacking elements of the second image, and drawing a volume cube of a stacking object based on a detection result to obtain an actual cube frame;

and calculating the volume ratio of the volume cube to the rated cube according to the pixel size data of the actual cube border and the rated cube border, and determining the product of the volume ratio and the rated square number as the volume detection value of the accumulation in the second image.

2. The detection method according to claim 1, wherein the steps of obtaining a first image of a target detection area in advance, labeling the first image, determining a rated cube border of the target detection area, and configuring a rated square number of the rated cube comprise:

aiming at the first image, manually marking each angular point of the ground range of a target detection area by adopting special marking software, and then integrally dragging and lifting to determine the angular point of the top surface, so that the drawing of the rated cube is realized, and the frame of the rated cube is determined;

and calculating a reference proportion value for determining the actual length of the fixed reference object in the target detection area and the pixel length of the fixed reference object in the first image, calculating a volume value of the rated cube based on the reference proportion value, and performing correlation processing to complete the configuration of the rated square number of the rated cube.

3. The detection method according to claim 1, wherein the performing stacked element detection on the second image and performing volume cube rendering of a stacked object based on the detection result to obtain an actual cube border comprises:

adopting a pre-selected trained segmentation algorithm model to infer and predict a mask region of the deposit in the second image;

view cutting is carried out on the mask area based on the coordinate and the trigonometric function of the rated cube frame, and a mask of the mask area in each view is obtained;

calculating and determining the length, width and height pixel values of the volume cube based on the area ratio of each mask in the corresponding view relative to the corresponding surface of the rated cube;

and drawing the volume cube according to the length, the width and the height pixel values of the volume cube to obtain the actual cube border.

4. The detection method according to claim 3, wherein the inference predicting the mask region of the deposit in the second image by using a pre-selected trained segmentation algorithm model specifically comprises:

performing clipping processing on the second image based on the rated cube border;

and carrying out reasoning and prediction on the cut image by adopting a pre-selected trained segmentation algorithm model to obtain the mask region.

5. The detection method according to claim 4, wherein the clipping processing is performed on the second image based on the rated cube border, specifically:

analyzing the coordinate values of each point on the rated cube border, and determining the extreme coordinate values min (x), min (y), max (x), max (y) in the coordinate values;

and (min (x), min (y)) and (max (x), max (y)) are used as diagonal points to construct a cutting rectangular frame, and the second image is cut according to the cutting rectangular frame.

6. The detection method according to claim 3, wherein the volume cube is drawn according to the length, width and height pixel values of the volume cube, specifically:

adjusting the length pixel value of the volume cube to be a pixel value of the bottom frame length of the rated cube, and drawing the volume cube according to the width pixel value, the height pixel value and the adjusted length pixel value of the volume cube;

in the drawing process, a vertex at the lower side of the rated cubic frame is used as a coordinate origin, and the directions of the frame line connected with the vertex in the volume cubic frame and the corresponding frame line in the rated cubic frame are kept consistent.

7. The detection method according to claim 3, wherein the segmentation algorithm model comprises a yolact algorithm model.

8. The inspection method of any one of claims 1 to 7, wherein the target inspection area comprises a loading/unloading port area of a logistics yard.

9. An apparatus for detecting the entire volume of a stack, comprising:

the system comprises a labeling processing module, a judging module and a judging module, wherein the labeling processing module is used for labeling a first image of a target detection area acquired in advance, determining a rated cube frame of the target detection area and configuring the rated square number of the rated cube;

the detection drawing processing module is used for detecting stacking elements of a second image and drawing a volume cube of a stacking object based on a detection result to obtain an actual cube frame, wherein the second image is obtained by shooting a camera position and parameters when the first image is obtained and shooting the target detection area in real time;

and the volume determination module is used for calculating the volume ratio of the volume cube to the rated cube according to the pixel size data of the actual cube border and the rated cube border, and determining the product of the volume ratio and the rated square number as the volume detection value of the accumulation in the second image.

10. An electronic device, comprising:

a memory having an executable program stored thereon;

a processor for executing the executable program in the memory to implement the steps of the method of any one of claims 1-8.

Technical Field

The application belongs to the technical field of logistics, and particularly relates to a method and a device for detecting the whole volume of a deposit and electronic equipment.

Background

With the development of express delivery in the logistics express delivery industry and the continuous updating of the technology of the internet of things, AI application scenes based on cameras can continuously fall to the ground in the logistics express delivery industry. More structured data related to services can be collected in a traditional logistics sorting field through an AI video technology of a camera, before supplement, the data are collected online intelligently and in real time through an offline manual collection and the like through a collection device camera and the like, and the services can be attached to the data base provided for an optimization process. Each big express delivery logistics company and the technique wave who experiences IOT become intelligent IOT equipment with ordinary camera, have realized that the digitization of station is visual, the intellectuality of management. The equipment cost is not increased, and the efficiency and the cost are obviously improved. The technology is used for driving cost reduction and efficiency improvement of express logistics.

In the related art, in the field of express delivery volume detection subdivision, currently, most of express deliveries adopt an industrial camera or a depth camera on a conveyor belt for volume detection. Such as the volume detection equipment supplied by shanghai real-dry technology. The device and the method can only detect the volume when a single express passes through the detection device. The detection of the entire volume of the pile up cannot be realized in a scene of a camera.

Still others are image-based 2D to 3D techniques that use depth learning algorithms to construct 3D models from multiple 2D images to detect volume. According to the scheme, multi-view sample data are difficult to acquire in a field station express accumulation scene, under the condition limitation, an operation field cannot use a camera with multiple machine positions in an accumulation area, and even if multiple machine positions can be set for shooting, the hardware cost is increased.

There are also logistics companies that try to do this with stacked image classification. The piled-up image is classified into 30%, 60% and 90%, and then the forced classification type is marked through data. The method has the problems that how to define the classification of the sample labels is, because the imaging effect and the size are different according to the difference of the distance of camera positions of the stations, different personnel can hardly define the classification by themselves when making the sample labels. Even if the model can be labeled by some artificial empirical constraints, the trained model can only be divided in a coarse granularity, and the number of the squares of the specific accumulation volume cannot be accurately obtained. And if the environment of the station is changed, the stack classification samples need to be marked again for training, namely the model cannot be generalized, the effect is not good when changing a sorting station, and the training needs to be carried out again.

The above is only for the purpose of assisting understanding of the technical aspects of the present invention, and does not represent an admission that the above is prior art.

Disclosure of Invention

For overcoming the problem that exists in the correlation technique at least to a certain extent, this application provides a deposit whole volume's detection method, device and electronic equipment, helps the whole volume detection of better realization piling up express delivery in the commodity circulation letter sorting place.

In order to achieve the purpose, the following technical scheme is adopted in the application:

in a first aspect,

the application provides a method for detecting the whole volume of a deposit, which comprises the following steps:

the method comprises the steps of obtaining a first image of a target detection area in advance, carrying out labeling processing on the first image, determining a rated cube frame of the target detection area, and configuring the rated square number of the rated cube;

acquiring a second image of the target detection area in real time based on the camera position and parameters when the first image is acquired, detecting stacking elements of the second image, and drawing a volume cube of a stacking object based on a detection result to obtain an actual cube frame;

and calculating the volume ratio of the volume cube to the rated cube according to the pixel size data of the actual cube border and the rated cube border, and determining the product of the volume ratio and the rated square number as the volume detection value of the accumulation in the second image.

Optionally, the obtaining a first image of a target detection region in advance and labeling the first image, determining a rated cube border of the target detection region, and configuring a rated square number of the rated cube includes:

aiming at the first image, manually marking each angular point of the ground range of a target detection area by adopting special marking software, and then integrally dragging and lifting to determine the angular point of the top surface, so that the drawing of the rated cube is realized, and the frame of the rated cube is determined;

and calculating a reference proportion value for determining the actual length of the fixed reference object in the target detection area and the pixel length of the fixed reference object in the first image, calculating a volume value of the rated cube based on the reference proportion value, and performing correlation processing to complete the configuration of the rated square number of the rated cube.

Optionally, the performing stacking element detection on the second image and performing volume cube drawing of a stack based on the detection result to obtain an actual cube frame includes:

adopting a pre-selected trained segmentation algorithm model to infer and predict a mask region of the deposit in the second image;

view cutting is carried out on the mask area based on the coordinate and the trigonometric function of the rated cube frame, and a mask of the mask area in each view is obtained;

calculating and determining the length, width and height pixel values of the volume cube based on the area ratio of each mask in the corresponding view relative to the corresponding surface of the rated cube;

and drawing the volume cube according to the length, the width and the height pixel values of the volume cube to obtain the actual cube border.

Optionally, the predicting a mask region of the deposit in the second image by inference using a pre-selected trained segmentation algorithm model specifically includes:

performing clipping processing on the second image based on the rated cube border;

and carrying out reasoning and prediction on the cut image by adopting a pre-selected trained segmentation algorithm model to obtain the mask region.

Optionally, for the second image, performing clipping processing based on the rated cube border specifically includes:

analyzing the coordinate values of each point on the rated cube border, and determining the extreme coordinate values min (x), min (y), max (x), max (y) in the coordinate values;

and (min (x), min (y)) and (max (x), max (y)) are used as diagonal points to construct a cutting rectangular frame, and the second image is cut according to the cutting rectangular frame.

Optionally, the volume cube is drawn according to the length, the width, and the height pixel values of the volume cube, specifically:

adjusting the length pixel value of the volume cube to be a pixel value of the bottom frame length of the rated cube, and drawing the volume cube according to the width pixel value, the height pixel value and the adjusted length pixel value of the volume cube;

in the drawing process, a vertex at the lower side of the rated cubic frame is used as a coordinate origin, and the directions of the frame line connected with the vertex in the volume cubic frame and the corresponding frame line in the rated cubic frame are kept consistent.

Optionally, the segmentation algorithm model comprises a yolact algorithm model.

Optionally, the target detection area comprises a loading/unloading port area of the logistics yard.

In a second aspect of the present invention,

the application provides a detection device of whole volume of deposit, and the device includes:

the system comprises a labeling processing module, a judging module and a judging module, wherein the labeling processing module is used for labeling a first image of a target detection area acquired in advance, determining a rated cube frame of the target detection area and configuring the rated square number of the rated cube;

the detection drawing processing module is used for detecting stacking elements of the second image and drawing a volume cube of a stacking object based on a detection result to obtain an actual cube frame, wherein the second image is obtained by shooting a camera position and parameters when the first image is obtained and shooting the target detection area in real time;

and the volume determination module is used for calculating the volume ratio of the volume cube to the rated cube according to the pixel size data of the actual cube border and the rated cube border, and determining the product of the volume ratio and the rated square number as the volume detection value of the accumulation in the second image.

In a third aspect,

the application provides an electronic device, including:

a memory having an executable program stored thereon;

a processor for executing the executable program in the memory to implement the steps of the method described above.

This application adopts above technical scheme, possesses following beneficial effect at least:

the method adopts an image visual volume construction technology, an actual stacking cube can be constructed in the rated cube as long as the rated cube and the rated square number are marked in a detection area, and the final actual stacking volume square number is calculated by the condition that the ratio of the constructed rated cube to the actual stacking cube is equal to the ratio of the rated square number to the actual square number. According to the method, no physical equipment is needed to be added, the volume detection can be carried out based on the single-path common camera installed in the logistics field station, the model involved in the method can be well generalized, a large number of samples are not needed to be marked, and the rapid deployment and the online are facilitated.

Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.

Drawings

The accompanying drawings are included to provide a further understanding of the technology or prior art of the present application and are incorporated in and constitute a part of this specification. The drawings expressing the embodiments of the present application are used for explaining the technical solutions of the present application, and should not be construed as limiting the technical solutions of the present application.

Fig. 1 is a schematic flow chart of a method for detecting the entire volume of a heap according to an embodiment of the present application;

fig. 2 is a schematic drawing illustrating a nominal cube in the method for detecting the entire volume of a heap according to an embodiment of the present application;

fig. 3 is a schematic explanatory view of an effect of arrangement of the rated number of squares of the rated cube in the method for detecting the entire volume of a deposit according to the embodiment of the present application;

fig. 4 is a schematic illustration of the inference prediction effect of the mask region in the method for detecting the entire volume of a deposit according to an embodiment of the present application;

fig. 5 is a schematic illustration of volume cube mapping in a method for detecting the entire volume of a heap according to an embodiment of the present application;

fig. 6 is a schematic explanatory view of the effect of determining a volume detection value of a deposit in the method for detecting the entire volume of a deposit according to the embodiment of the present application;

FIG. 7 is a schematic illustration of the application of the method for detecting the entire volume of a heap to a plurality of detection regions according to one embodiment of the present application;

FIG. 8 is a schematic structural view of a device for detecting the entire volume of a heap according to an embodiment of the present application;

fig. 9 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Detailed Description

In order to make the objects, technical solutions and advantages of the present application more apparent, the technical solutions of the present application will be described in detail below. It is to be understood that the embodiments described are only a few embodiments of the present application and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the examples given herein without making any creative effort, shall fall within the protection scope of the present application.

As described in the background art, the prior art related to express delivery volume detection in practice has various defects, and in view of the above, the application provides a method for detecting the overall volume of a deposit, which is beneficial to better realizing the overall volume detection of the deposited express delivery in a logistics sorting field.

In one embodiment, as shown in fig. 1, the method for detecting the whole volume of a heap in the present application comprises the following steps:

step S110, a first image of a target detection area is obtained in advance and labeled, a nominal cube frame of the target detection area is determined, and a nominal square number of the nominal cube is configured.

The application scenario of the embodiment of the application is a logistics sorting station, the target detection area is a loading and unloading port area of the logistics station, the loading and unloading port in the application means that the position where each logistics vehicle stops in the station is defined as one loading and unloading port, each loading and unloading port is provided with one conveyor belt for conveying goods, express goods can be stacked on two sides of each conveyor belt, and the areas on two sides belong to the target detection area.

In the embodiment, a first image is obtained by capturing an image through a camera configured on a collection station; specifically, in step S110, for the first image, a special annotation software is used to manually annotate each corner of the ground range of the target detection area, and then the whole is dragged and lifted to determine the top corner, so as to draw the rated cube and determine the frame of the rated cube. For example, as shown in fig. 1, only 4 points at the bottom are marked, and then the whole body is dragged and lifted, so that 4 points on the upper top surface can be automatically determined, and the drawing of 8 points of the cube is completed.

In step S110, in order to configure the rated square number, a reference ratio value between the actual length of the fixed reference object in the target detection region and the pixel length of the fixed reference object in the first image is determined by calculation, and a volume value of the rated cube is calculated based on the reference ratio value and is subjected to correlation processing, so as to complete the configuration of the rated square number of the rated cube;

for example, the loading/unloading port is respectively provided with an area A and an area B corresponding to two sides of the conveyor belt. And respectively carrying out image cube labeling and recording of rated volume (rated square number). Rated volume is length x width x height. When in recording, the data are directly recorded according to the length of the parameter proportion, such as: the length of the carriage (reference object) is 2.5, the length, width and height of the rated volume are recorded by referring to the fixed length proportion of the periphery, and the configuration result is shown in figure 3.

The method processing step of step S110 is also referred to as basic parameter information configuration in this application.

Then, step S120 is performed to obtain a second image of the target detection area in real time based on the camera position and parameters obtained when the first image is obtained, perform stacking element detection on the second image, and perform volume cube drawing of a stack based on the detection result, so as to obtain an actual cube frame.

In particular, the method comprises the following steps of,

in step S120, a pre-selected trained segmentation algorithm model is first used to predict a mask region of a deposit in the second image by inference, i.e., to perform detection of a deposit element.

In order to efficiently obtain the mask area of the deposit, the segmentation algorithm model can adopt a yolact algorithm model, for example, as shown in fig. 4, the area enveloped by the outer contour line of the express deposit area in the right side of fig. 4 is the mask area of the presumably detected deposit;

the yolcat model is a deep learning algorithm model, has higher inference speed (fps >30) compared with a common maskrnn model, and adopts a simple full-convolution model to divide an instance segmentation task into two parallel subtasks: a set of prototype (prototype) masks is generated and mask coefficients are predicted for each instance. Finally, the prototype and the template coefficients are linearly combined to generate instance masks. In practical deployment applications, yolcat requires only few samples to converge the model quickly, since the entire process is independent of repooling, thereby yielding very high quality masks. After deployment, a good reasoning effect can be achieved. Specifically, in the model training phase, sample data is converted into a data set in a coco format, and training of the yolact depth model is performed.

Preferably, in the model application, in order to improve the efficiency (reduce the number of matrix calculations and increase the speed of inference), the clipping process can be performed on the second image based on the rated cube frame; and then, carrying out reasoning and prediction on the cut image by adopting a pre-selected trained segmentation algorithm model to obtain a mask region. Similarly, to improve efficiency, similar pre-processing may be applied to the sample data during the model training phase.

In the above process, the clipping processing mode is to analyze the coordinate values of each point on the rated cube frame, and determine the extreme coordinate values min (x), min (y), max (x), and max (y); and (min (x), min (y)) and (max (x), max (y)) are used as the diagonal points to construct a cutting rectangular frame, and the second image is cut according to the cutting rectangular frame.

In step S120, after the mask area is obtained through inference and prediction, view cutting is performed on the mask area based on the coordinate of the rated cube frame and the trigonometric function to obtain a mask of the mask area in each view;

the view cutting can make the mask calculation of the whole stacking package fall into four views, so that the purpose of cutting is to calculate the length, width and height of the stacking actual area, each view plane represents a dimension calculation, for example, the left view and the right view and the bottom view represent the calculation of the index of the width, the front view represents the calculation of the index of the height, and the like.

In step S120, after the masks of the mask regions in the views are obtained, the length, width, and height pixel values of the volume cube are determined by calculating, based on the area ratio of each mask in the corresponding view with respect to the corresponding surface of the rated cube, that is, performing contour extraction and then area calculation on the segmented mask image, calculating the ratio of length to width using the mask area ratio of each surface (with respect to the corresponding surface of the rated cube), and determining the length, width, and height pixel values of the volume cube.

In step S120, the volume cube is finally drawn according to the length, width, and height pixel values of the volume cube, so as to obtain an actual cube border.

It should be noted that, in practice, since express deposits may exceed the outside of the front view of the rated cube, in order to facilitate subsequent processing, when the volume cube is drawn, the length pixel value of the volume cube may be adjusted to the pixel value of the bottom frame length of the rated cube, and the volume cube is drawn according to the width pixel, the height pixel value and the adjusted length pixel value of the volume cube; in the drawing process, a vertex at the lower side of the rated cubic frame is used as a coordinate origin, and the directions of a frame line connected with the vertex in the volume cubic frame and a corresponding frame line in the rated cubic frame are kept consistent. In other words, the final rendered visual effect is as shown in fig. 5, with the volume cube completely within the nominal cube and aligned.

In addition, in some special cases, due to less express delivery stacking, when view cutting and calculation are performed to determine the length, width and height pixel values of the volume cube, the height pixel value cannot be obtained, and at this time, the height pixel value is determined to be a preset default value, for example, the preset default value is 1 pixel.

In the method processing step of step S120, the step following the detection of the stacking elements is also referred to as the synthesis of the image cube volume in the present application.

Returning to fig. 1, finally, in step S130, the volume ratio of the volume cube to the rated cube is calculated according to the pixel size data of the actual cube border and the rated cube border, and the product of the volume ratio and the rated square number is determined as the volume detection value of the stacked object in the second image.

In this application, for the calculation of the actual stacking volume of each target detection region, we define a formula: image actual volume/image nominal volume-actual packing volume/nominal packing volume,

the image rated volume is the product of pixel values of the displayed length, width and height when a rated cube is drawn;

the rated stacking volume is the rated square number which is correspondingly configured after a rated cube is drawn;

the actual volume of the image is a product of pixel values of the length, width and height of the volume cube calculated after the image cube is synthesized.

Based on the formula, it is obvious that the actual stacking volume is the nominal stacking volume (image actual volume/image nominal volume), that is, in step S130, as mentioned, the volume ratio (pixel volume compared) of the volume cube and the nominal cube is calculated, and the product of the volume ratio and the nominal square is determined as the detected volume value of the stacking object in the second image, for example, the prediction result is as shown in fig. 6.

In addition, as shown in fig. 7, in an actual scene, there are a plurality of loading and unloading ports in the logistics site, and the plurality of loading and unloading ports have a plurality of region stacking volumes, and the final station stacking volume calculation actually includes synthesizing each region image on an original image to perform an estimation calculation, and summing the volumes of each region. Specifically, in practice, the volumes of the corresponding areas can be aggregated by the dimension of the loading and unloading ports to form the stacking volume corresponding to each loading and unloading port, and finally the stacking volumes of each loading and unloading port are integrally summed to form the stacking volume of the loading and unloading port of the whole station, so that the method is applied to stacking alarm, keeping volume statistics and the like.

The method adopts an image visual volume construction technology, an actual stacking cube can be constructed in the rated cube as long as the rated cube and the rated square number are marked in a detection area, and the final actual stacking volume square number is calculated by the condition that the ratio of the constructed rated cube to the actual stacking cube is equal to the ratio of the rated square number to the actual square number. According to the method, no physical equipment is needed to be added, the volume detection can be carried out based on the single-path common camera installed in the logistics field station, the model involved in the method can be well generalized, a large number of samples are not needed to be marked, and the rapid deployment and the online are facilitated.

The technical act in the present application can be used to implement stack warning function in logistics yard management, for example, a loading/unloading port of a sorting yard finds more than 2 sides of stack, but no vehicle is loaded and a warning is given. And reminding a dispatcher to arrange the vehicle to come for operation as soon as possible. The system can also be used for realizing the functions of warehouse remaining inspection and route scheduling in station management, for example, the logistics sorting field is in the early morning, the loading and unloading port accumulation volume acquired by the camera is integrated with the route data, and the route scheduling is convenient to arrange on the next day.

Fig. 8 is a schematic structural diagram of a device for detecting the entire volume of a heap according to an embodiment of the present invention, and as shown in fig. 8, the device 800 includes:

a labeling processing module 801, configured to label a first image of a target detection area acquired in advance, determine a rated cube border of the target detection area, and configure a rated square number of the rated cube;

a detection drawing processing module 802, configured to perform stacking element detection on a second image and perform volume cube drawing of a stack based on a detection result to obtain an actual cube frame, where the second image is an image obtained by shooting a camera position and parameters when the first image is obtained and shooting a target detection area in real time;

and a volume determining module 803, configured to calculate a volume ratio of the volume cube to the rated cube according to the pixel size data of the actual cube border and the rated cube border, and determine a product of the volume ratio and the rated square number as a volume detection value of the stack in the second image.

With respect to the detection apparatus 800 in the above related embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment related to the method, and will not be elaborated here.

Fig. 9 is a schematic structural diagram of an electronic device according to an embodiment of the present application, and as shown in fig. 9, the electronic device 900 includes:

a memory 901 on which an executable program is stored;

a processor 902 for executing the executable program in the memory 901 to implement the steps of the above method.

With respect to the electronic device 900 in the above-mentioned embodiment, the specific manner of executing the program in the memory 901 by the processor 902 thereof has been described in detail in the embodiment related to the method, and will not be elaborated herein.

The above description is only a preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.