Dimension measurement device and method based on learning model

Abstract

The size measurement apparatus and method according to an embodiment of the present invention can acquire size information of a specific object in an image of interest by using a pre-trained learning model, thereby obtaining highly reliable size measurement result data with enhanced accuracy.

Description

Technical Field

[0001] This application claims priority to and the benefit of Korean Patent Application No. 10-2022-0109769 filed in the Korean Intellectual Property Office on August 31, 2022, and Korean Patent Application No. 10-2023-0103918 filed in the Korean Intellectual Property Office on August 9, 2023, the entire contents of which are incorporated herein by reference.

[0002] The present invention relates to an apparatus and method for measuring size, and more particularly, to an apparatus and method for acquiring size information of a measurement object in an image of interest using a pre-trained learning model. Background Art

[0003] As the price of energy rises due to the depletion of fossil fuels and interest in environmental pollution increases, the demand for secondary batteries as an environmentally friendly alternative energy source is rapidly increasing.

[0004] Lithium secondary batteries among secondary batteries are being used in various industrial fields from mobile application devices to automobiles, robots, and energy storage devices in response to today's environmental regulations and high oil prices.

[0005] These lithium batteries are generally classified into a cylindrical, square, or pouch type according to the shape of an exterior material in which an electrode assembly is accommodated.

[0006] Among them, the cylindrical battery can be provided in a CTP structure consisting of a plurality of battery cells. In other words, the cylindrical battery is provided in a structure in which an electrode including a separator between a negative electrode and a positive electrode is wound and inserted into a battery can for assembly.

[0007] These cylindrical batteries undergo quality inspection after the assembly process.

[0008] Typically, in quality inspection of cylindrical batteries, an image of the object to be measured is used to measure the size of the relevant area in order to determine the location of the weld, the tension of the weld, or any defects in the battery components that occurred during any manufacturing process.

[0009] A conventional size measuring apparatus detects an edge of a measuring object corresponding to a preset threshold brightness value from an object image as a brightness image (grayscale model), extracts pixel information of the detected edge, and obtains size information of the measuring object.

[0010] However, the conventional dimension measuring apparatus has a disadvantage that when an error occurs in program logic pre-designed by an engineer, a boundary line of a measurement object on an object image is erroneously detected, thereby reducing the reliability of dimension information. Summary of the invention

[0011] Technical issues

[0012] To avoid one or more problems in the prior art, embodiments of the present disclosure provide an apparatus for measuring size.

[0013] In order to avoid one or more problems in the prior art, an embodiment of the present disclosure further provides a method for measuring size.

[0014] Technical Solution

[0015] To achieve the purpose of the present disclosure, a dimension measurement device based on a learning model may include: at least one processor and a memory, the memory being configured to store at least one instruction executed by the at least one processor, and the at least one instruction may include: an instruction to obtain an image of interest containing a measurement object whose size is to be measured from an object image; and an instruction to input the image of interest into a pre-trained learning model and output the size information of the measurement object as result data.

[0016] Here, the pre-trained learning model may be a learning model based on deep learning.

[0017] The instruction to output the size information of the measured object as result data may include an instruction to segment the image of interest into at least one physical object using a pre-trained learning model and mark size measurement points of the measured object.

[0018] The instructions to acquire the image of interest may include instructions to obtain the image of interest including the measurement object from the object image using a rule-based algorithm.

[0019] The instruction to acquire the image of interest may include an instruction to obtain the image of interest by performing image preprocessing on the object image.

[0020] According to another embodiment of the present disclosure, a method for dimension measurement based on a learning model may include the following steps: acquiring an image of interest containing a measurement object whose size is to be measured from an object image; and inputting the image of interest into a pre-trained learning model and outputting size information of the measurement object as result data.

[0021] Here, the pre-trained learning model may be a learning model based on deep learning.

[0022] The step of outputting the size information of the measured object as result data may include segmenting the image of interest into at least one physical object using a pre-trained learning model and marking size measurement points of the measured object.

[0023] The step of acquiring the image of interest may include obtaining the image of interest including the measurement object from the object image using a rule-based algorithm.

[0024] The step of acquiring the image of interest may include obtaining the image of interest by performing image preprocessing on the object image.

[0025] Beneficial Effects

[0026] The size measurement apparatus and method according to the embodiments of the present invention can obtain size information of a specific object in an image of interest by using a pre-trained learning model, thereby obtaining highly reliable size measurement result data with improved accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 An erroneous detection image of a welded component measured by a general dimension measuring device is shown.

[0028] Figure 2 is a block diagram of an apparatus for measuring a size according to an embodiment of the present invention.

[0029] Figure 3 is a flowchart of a dimension measuring method using the dimension measuring apparatus according to an embodiment of the present invention.

[0030] Figure 4 : is an image of an image of interest for explaining the dimension measuring apparatus according to the embodiment of the present invention.

[0031] Figure 5 are an image of the first current collector plate acquired by the dimension measuring apparatus according to an embodiment of the present invention and an image of interest of a welding region extracted from the image of the first current collector plate.

[0032] Figure 6 The method for extracting Figure 5 Preprocessed images of the image of interest in .

[0033] Figure 7 An image of a second current collector plate according to another embodiment of the present invention and an image of interest extracted from the image of the second current collector plate are shown.

[0034] Figure 8 Another embodiment of the present invention is shown for extracting Figure 7 Preprocessed images of the image of interest in .

[0035] Fig. 9 An image for explaining a method of measuring a size of an electrode region using a pre-trained learning model in a size measuring apparatus according to another embodiment of the present invention is shown.

[0036] 100: Memory 200: Processor

[0037] 300: transceiver 400: input interface

[0038] 500: output interface 600: storage device

[0039] 700: Bus DETAILED DESCRIPTION

[0040] The present invention can be modified in various forms and has various embodiments, and its specific embodiments are shown in the accompanying drawings by way of example and will be described in detail below. However, it should be understood that it is not intended to limit the present invention to specific embodiments, but on the contrary, the present invention will cover all modifications, equivalents and alternatives that fall within the spirit and technical scope of the present invention. In the description of the entire drawings, the same reference numerals refer to the same elements.

[0041] It will be understood that, although terms such as first, second, A, B, etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another element. For example, without departing from the scope of the present invention, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element. As used herein, the term "and / or" includes a combination of any of a plurality of associated listings or a plurality of associated listings.

[0042] It will be understood that when an element is referred to as being "coupled" or "connected" to another element, it can be directly coupled or connected to the other element, or there may be intervening elements. Conversely, when an element is referred to as being "directly coupled" or "directly connected" to another element, there are no intervening elements.

[0043] The terms used herein are only used for the purpose of describing specific embodiments and are not intended to limit the present invention. As used herein, unless the context clearly indicates otherwise, the singular forms "one", "an" and "the" are also intended to include plural forms. It will also be understood that the terms "comprise", "contain", "include", "includes" and / or "have" when used herein specify the presence of the features, integers, steps, operations, constituent elements, parts and / or combinations thereof, but do not exclude the presence or addition of one or more other features, integers, steps, operations, constituent elements, parts and / or combinations thereof.

[0044] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as those commonly understood by those skilled in the art to which the present invention belongs. It will also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant art, and will not be interpreted in an idealized or overly formal sense unless explicitly defined as such herein.

[0045] Figure 1 An erroneous detection image of a welded component measured by a general dimension measuring device is shown.

[0046] Reference Figure 1 , general dimensional measuring equipment checks the quality of manufactured batteries after the assembly process.

[0047] Generally, when inspecting the quality of a battery, dimensional information on at least one measurement object is required to inspect the position and tension of a welded component, or to determine whether the battery component has any defects occurring in a manufacturing process.

[0048] Therefore, a general size measuring device acquires an object image including a measurement object, applies a pre-designed program logic to the object image to obtain the edge of the measurement object, and then measures the size of the measurement object based on pixel information of the boundary line.

[0049] For example, a general dimension measuring device applies a pre-designed program logic to a brightness image (grayscale model) in which the measuring object is photographed. Therefore, the dimension measuring device detects at least one pixel corresponding to a preset detection condition value (e.g., a threshold brightness value), and obtains a boundary line of the measuring object based on this. Thereafter, the dimension measuring device measures the size of the measuring object based on the pixel information of the boundary line. Here, the pre-designed program logic can be a rule-based algorithm model. In addition, the information about at least one pixel can include at least one of the height, width, peak white, peak dark, and aspect ratio of the at least one pixel.

[0050] In addition, the program logic applied to the general dimension measuring device may be pre-designed based on the previously secured object image. Therefore, when the brightness of the object image changes or the position of the measuring object within the object image changes, the general dimension measuring device to which the pre-designed program logic is applied may erroneously detect the boundary line of the measuring object or may incorrectly specify the measurement range of the boundary line, which leads to the disadvantage of reducing the reliability of the measured dimension information.

[0051] Therefore, in order to solve this problem, the present invention proposes a dimension measurement device and method with improved reliability, which uses a pre-trained machine learning model.

[0052] Hereinafter, preferred embodiments according to the present invention will be described in detail with reference to the accompanying drawings.

[0053] Figure 2 is a block diagram of an apparatus for measuring a size according to an embodiment of the present invention.

[0054] Reference Figure 2 The dimension measuring device according to the embodiment of the present invention can be used during the quality inspection performed after the assembly process of the cylindrical battery. However, it is not limited to the described content and can be applied to any process requiring dimension measurement.

[0055] The size measuring device can obtain size information of the measured object by inputting the object image into a pre-trained learning model. Here, the object image can be an image of at least a portion of the assembled battery.

[0056] More specifically, the dimension measuring apparatus may acquire an object image including a measurement target, and extract an image of interest including a region of interest (ROI) of the measurement target from the object image.

[0057] Thereafter, the size measuring device can obtain the size information of the measured object by inputting the image of interest into the pre-trained learning model.

[0058] Describing the size measuring apparatus in detail by configuration, the size measuring apparatus may include a memory 100 , a processor 200 , a transceiver 300 , an input interface 400 , an output interface 500 , and a storage device 600 .

[0059] According to the embodiment, the respective components 100 , 200 , 300 , 400 , 500 , and 600 included in the dimension measuring apparatus are connected through a bus 700 and can communicate with each other.

[0060] Among these components 100, 200, 300, 400, 500 and 600, the memory 100 and the storage device 600 may include at least one of a volatile storage medium and a non-volatile storage medium. For example, the memory 100 and the storage device 600 may include at least one of a read-only memory (ROM) and a random access memory (RAM).

[0061] The memory 100 may include at least one instruction executed by the processor 200 .

[0062] According to an embodiment, at least one instruction may include: an instruction for acquiring an image of interest including a measurement object whose size is to be measured from an object image; and an instruction for inputting the image of interest into a pre-trained learning model and outputting size information of the measurement object as result data.

[0063] The pre-trained learning model may be a deep learning based learning model.

[0064] The instructions for outputting the size information of the measured object as result data include instructions for segmenting the image of interest into at least one physical object using a pre-trained learning model and marking size measurement points of the measured object.

[0065] The instructions to acquire the image of interest may include instructions to obtain the image of interest including the measurement object from the object image using a rule-based algorithm.

[0066] The instruction to acquire the image of interest may include an instruction to obtain the image of interest by performing image preprocessing on the object image.

[0067] In addition, the processor 200 may refer to a central processing unit (CPU), a graphics processing unit (GPU), or a dedicated processor that executes the method according to the embodiment of the present invention.

[0068] As described above, the processor 200 may execute at least one program command stored in the memory 100 .

[0069] Figure 3 is a flowchart of a dimension measuring method using the dimension measuring apparatus according to an embodiment of the present invention.

[0070] Reference Figure 3 , the size measuring device according to the embodiment of the present invention may acquire an object image (S1000). Here, the object image may be an image capturing at least a portion of a fully assembled battery, and the object image may be an image including a measurement object for size measurement.

[0071] According to one embodiment, the object image may be a video image of at least one welding area as a measurement target. For example, the welding area may be a welding area formed on a first current collector or a second current collector in a battery. Here, the first current collector and the second current collector may be a negative electrode current collector and a positive electrode current collector, respectively.

[0072] According to another embodiment, the object image may be an X-ray image in which at least one electrode region as a measurement target is photographed. For example, the electrode region may include a first electrode or a second electrode of a battery.

[0073] Figure 4 : is an image of an image of interest for explaining the dimension measuring apparatus according to the embodiment of the present invention.

[0074] Reference Figure 3 and Figure 4, the dimension measuring device may acquire an image of interest from the object image (S3000). Here, the image of interest may be a partial image in which a region of interest (ROI) of a measurement target is partially extracted from the object image. According to one embodiment, the dimension measuring device may extract the image of interest after preprocessing the object image using a rule-based algorithm.

[0075] Thereafter, the dimension measuring device may obtain dimension information about the measured object from the image of interest using a pre-trained learning model (S5000). In other words, the dimension measuring device may obtain dimension information about the measured object by inputting the image of interest into the pre-trained learning model. Here, the machine learning model may be a classification learning model using deep learning.

[0076] To describe in more detail according to the embodiment, the dimension measuring device can use a classification learning model using deep learning that is pre-trained by training data to segment the image of interest as input data (features) into measurement objects (categories) to be measured. Thereafter, the dimension measuring device can mark the dimension measurement points of the measurement object, measure the pixel distance between the two marked points, and output it as result data.

[0077] Figure 5 and Figure 6 is an image for explaining a method of measuring the size of a welding area using an apparatus for measuring a size according to an embodiment of the present invention. More specifically, Figure 5 is an image of the first current collector plate and an image of interest of the welding region extracted from the image of the first current collector plate, and Figure 6 Shows the extraction Figure 5 The preprocessed image of the image of interest.

[0078] Reference Figure 5 , the dimension measuring device according to an embodiment of the present invention can obtain an image of the first current collector plate, which is an object image of a fully assembled battery as described above. For example, the first current collector plate can be a negative electrode current collector plate.

[0079] Thereafter, the size measuring apparatus may obtain an image of interest by preprocessing the image of the first current collector plate as the object image. Here, the image of interest may be an image including a welding area which is a region of interest (ROI) in the image of the first current collector plate.

[0080] For example, when the image of the first current collector plate includes a plurality of welding regions, the size measuring apparatus may separately extract a plurality of images of interest, each image of interest including each welding region.

[0081] For a more detailed description of the method for extracting images of interest, refer to Figure 6, the dimension measuring device may pre-process the image of the first current collector plate to detect at least one position of the at least one welding region from the image of the first current collector plate.

[0082] For example, in the case of a cylindrical battery whose assembly process has been completed, a plurality of electrode tabs may be provided by the electrode wound inside the can in a multilayer state in contact with the region on the first current collector plate. Here, at least one welding region may be formed on the first current collector plate for electrical contact with the electrode tab. In addition, at least one welding region may be formed on the first current collector plate in a diagonal direction (see Figure 5 ). Therefore, the dimension measuring device may perform a pre-processing of rotating the first current collector image so that the welding area is positioned horizontally or vertically.

[0083] Thereafter, the size measuring device may identify the location of the weld region as a region of interest based on the rotation image of the first current collector plate.

[0084] To describe in more detail according to the embodiment, the size measuring device may convert the rotation image of the first current collector plate into a gray value image according to a preset rule-based algorithm. Thereafter, the size measuring device may obtain the edge of the welding area in the image of the first current collector by inputting a predetermined condition value. For example, the predetermined condition value may be a preset pixel brightness value. In other words, the size measuring device may extract at least one pixel having a preset pixel brightness value as the boundary of the welding area. Thereafter, the size measuring device may identify the position of the welding area based on at least one pixel information extracted as the boundary of the welding area.

[0085] Thereafter, the size measuring device may convert the grayscale image of the first current collector plate of the value into a binary image. Then, as Figure 6 As shown, the dimension measuring device can obtain the starting coordinate point and the ending coordinate point (starting point, end point) of the welding area.

[0086] According to one embodiment, when the welding area in the image of the first current collector plate is located in the horizontal direction, the starting coordinate point may be a pixel coordinate value located at the left edge of the welding area relative to the x-axis, and the ending coordinate point may be a pixel coordinate value located at the right edge of the welding area relative to the x-axis.

[0087] According to another embodiment, when the welding area in the image of the first current collector plate is located in the vertical direction, the starting coordinate point may be the pixel coordinate value of the welding portion located at the top relative to the y-axis, and the ending coordinate point may be the pixel coordinate value located at the bottom relative to the y-axis.

[0088] According to an embodiment, the size measuring device may obtain the starting coordinate point and the ending coordinate point (starting point, end point) within the image of the first current collector plate based on a predetermined pixel brightness value. For example, when the welding area in the image of the first current collector plate is located in the horizontal direction, the size measuring device may check the pixel brightness value of each coordinate in the x-axis direction and compare it with the preset pixel brightness value. Here, the size measuring device may set the first coordinate point having a brightness value greater than or less than the preset pixel brightness value as the starting coordinate point (starting point). In addition, the size measuring device may obtain the last coordinate point having a brightness value greater than or less than the preset pixel brightness value as the end point.

[0089] In addition, when the welding area in the image of the first current collector plate is located in the vertical direction, the size measuring device can check the pixel brightness value of each coordinate in the y-axis direction and compare it with the preset pixel brightness value. Here, the size measuring device can set the first coordinate point having a brightness value greater than or less than the preset pixel brightness value as the starting point. In addition, the size measuring device can obtain the last coordinate point having a brightness value greater than or less than the preset pixel brightness value as the end point.

[0090] Thereafter, the size measuring device may acquire at least one size measurement point based on the extracted boundary information of the welding area and the starting coordinate point and the ending coordinate point (starting point, end point). For example, the size measuring device may use a sub-pixel algorithm to obtain at least one size measurement point of the welding area. Here, the sub-pixel algorithm may be a method of determining a point with a brightness value of 1 pixel or less as a corner of an object using an interpolation method. Therefore, the size measuring device may ultimately obtain the boundary of the welding area in the image of the first collector plate based on the information of at least one size measurement point.

[0091] Therefore, the dimension measuring device according to the embodiment of the present invention can extract an image of interest from the image of the first current collector plate based on the boundary information and dimension measurement point information of the welding area finally obtained through the pre-processing process. Thereafter, the extracted image of interest can be image processed (superimposed) by the dimension measuring device.

[0092] Thereafter, the dimensional measurement device can input the processed image of interest into a pre-trained deep learning-based classification learning model.

[0093] Thereafter, the size measurement device may segment the image of interest into physical units (i.e., the welding area of ​​the first current collector plate) using a pre-trained classification learning model, and then mark the size measurement points of the welding area. For example, the size measurement device may classify the objects in the image of interest in units of pixels, and then mark the size measurement points of the welding area.

[0094] Then, the dimension measuring device can measure the pixel distance between the two marked dimension measurement points based on the pre-trained deep learning-based classification learning model, thereby obtaining the dimension of the measured object as result data. Here, the classification learning model can be a model that transforms at least one training data into various forms and then uses it as training data in order to increase the amount of training data. For example, the training data can be transformed in various ways, such as turning it upside down or flipping it left and right, changing the brightness, saturation or hue, changing the contrast, rotating or moving it, and adding noise.

[0095] Figure 7 and Figure 8 is an image for explaining a method of measuring the size of a welding area using a size measuring apparatus according to another embodiment of the present invention. More specifically, Figure 7 An image of the second current collector plate and an image of interest extracted from the image of the second current collector plate are shown, and Figure 8 shows the preprocessing to extract Figure 7 The image of interest in .

[0096] Reference Figure 7 and Figure 8 As described above, the dimension measuring device according to another embodiment of the present invention can obtain an image of the second current collector plate as an object image. Here, the image of the second current collector plate is an image taken after the battery assembly process, and can be an image of the positive electrode current collector plate.

[0097] Thereafter, the size measuring device may perform image preprocessing based on the image of the second current collector plate to obtain an image of interest.

[0098] To explain more specifically, generally, the welding area on the image of the second current collector may have a specific pattern shape arranged symmetrically at the origin, such as Figure 7 Therefore, the dimension measuring device may transform the object image into polar space data so that the welding area, which is the area of ​​interest in the image of the second current collector plate, is horizontally located.

[0099] Thereafter, the size measuring device may identify the location of the weld region as a region of interest based on the transformed image of the second current collector plate.

[0100] To describe in more detail according to an embodiment, the size measuring device may convert the rotation image of the second collector plate into a gray value image according to a preset rule-based algorithm. Thereafter, the size measuring device may input a predetermined condition value to obtain the edge of the welding area in the image of the second collector plate.

[0101] Thereafter, the size measuring device may convert the grayscale image of the second current collector plate into a binary image. Then, the size measuring device may obtain the starting coordinate point and the ending coordinate point (starting point, end point) of the welding area. According to one embodiment, the size measuring device may obtain the starting coordinate point and the ending coordinate point (starting point, end point) within the image of the second current collector plate based on a predetermined pixel brightness value.

[0102] Thereafter, the dimension measuring device can obtain at least one dimension measurement point based on the extracted boundary information of the welding area and the starting coordinate point and the ending coordinate point (starting point, end point). Therefore, the dimension measuring device can extract an image of interest from the image of the second current collector plate based on the welding dimension measurement point information and the boundary information of the welding area obtained by the preprocessing process.

[0103] Thereafter, the extracted image of interest may be subjected to image processing (superposition) by the size measuring device.

[0104] Thereafter, the dimension measuring device may input the image of interest after image processing into a pre-trained deep learning-based classification learning model.

[0105] Thereafter, the size measuring device may segment the image of interest into at least one welding region of the second current collector plate as a physical unit using a pre-trained classification learning model.

[0106] Thereafter, the dimensional measurement device can use a pre-trained deep learning-based classification learning model (deep learning) to mark the dimensional measurement points of the welding area.

[0107] Thereafter, the size measuring device can obtain the size of the measured object as result data by measuring the pixel distance between the two marked size measurement points. Here, the classification learning model can be a model that transforms at least one training data into various forms and then uses it as training data in order to increase the amount of training data. For example, the training data can be transformed in various ways, such as turning it upside down or flipping it left and right, changing the brightness, saturation or hue, changing the contrast, rotating or moving it, and adding noise.

[0108] Fig. 9 An image for explaining a method of measuring a size of an electrode region using a pre-trained learning model in a size measuring apparatus according to another embodiment of the present invention is shown.

[0109] Reference Fig. 9 As described above, the dimension measuring apparatus according to another embodiment of the present invention can obtain an object image from an image of an electrode region captured after an assembly process. For example, the electrode image can be an X-ray image.

[0110] Thereafter, the size measuring device may extract an image of interest including a connection region of the negative electrode and the positive electrode, which is a region of interest (ROI), from the electrode region image.

[0111] Thereafter, the extracted image of interest may be subjected to image processing (superposition) by the size measuring device.

[0112] Thereafter, the size measuring device may input the image of interest of the processed image into a pre-trained classification learning model based on deep learning. Thus, the size measuring device may obtain the length data of the electrode region connecting the positive terminal and the negative terminal of the electrode.

[0113] More specifically, the size measurement device can segment the image of interest input into the classification learning model into physical units according to a pre-trained classification learning model based on deep learning. According to one embodiment, the classification learning model can be a semantic segmentation model. For example, the size measurement device can segment the image of interest into electrode images.

[0114] Thereafter, the dimension measuring device may mark the dimension measuring points of the electrode region based on the classification learning model. For example, the dimension measuring points within the electrode region are the positive end of the electrode located near the joint region (the first point), and the negative end of the electrode located near the electrode body (the region where the coating region of the electrode is formed) (the second point). Here, the dimension measuring device may mark the dimension measuring points in a plurality of numbers to correspond to the number of turns of the electrode.

[0115] Thereafter, the dimension measuring device can measure the pixel distance between the first and second points of the mark based on a pre-trained classification learning model based on deep learning, thereby obtaining the length of the electrode to be measured as result data. Here, the classification learning model can be a model that transforms at least one training data into various forms and then uses it as training data to increase the amount of training data. For example, the training data can be transformed in various ways, such as turning it upside down or flipping it left to right, changing the brightness, saturation or hue, changing the contrast, rotating or moving it, and adding noise.

[0116] The dimension measuring apparatus and method according to the embodiments of the present invention have been described above.

[0117] The size measurement apparatus and method according to the embodiments of the present invention can obtain size information of a specific object in an image of interest by using a pre-trained learning model, thereby obtaining highly reliable size measurement result data with improved accuracy.

[0118] The operation of the method according to an embodiment of the present invention can be implemented as a computer-readable program or code on a computer-readable recording medium. The computer-readable recording medium includes all types of recording devices having data readable by a computer system stored therein. In addition, the computer-readable recording medium can be distributed in a network-connected computer system to store and execute a computer-readable program or code in a distributed manner.

[0119] In addition, the computer readable recording medium may include hardware devices specially configured to store and execute program instructions such as ROM, RAM, flash memory, etc. The program instructions may include not only machine language codes created by a compiler, but also high-level language codes that can be executed by a computer using an interpreter.

[0120] Although some aspects of the present invention have been described in the context of equipment, it can also be represented according to the description of the corresponding method, wherein a piece or equipment corresponds to a method step or the feature of a method step. Similarly, the aspect described in the context of the method can also represent the feature of a corresponding piece or project or a corresponding device. Some or all of the method steps can be performed by (or using) hardware devices such as, for example, a microprocessor, a programmable computer or an electronic circuit. In some embodiments, one or more of the most important method steps can be performed by such equipment.

[0121] In the foregoing, the present invention has been described with reference to exemplary embodiments thereof, but those skilled in the art will appreciate that various corrections and changes may be made to the present invention within the scope without departing from the spirit and scope of the invention described in the appended claims.

Claims

1. A dimension measuring device based on a learning model, the dimension measuring device comprising: at least one processor; as well as a memory configured to store at least one instruction executed by the at least one processor; The at least one instruction includes: An instruction for acquiring an image of interest including a measurement object whose size is to be measured from the object image; and Instructions for inputting the image of interest into a pre-trained learning model and outputting size information of the measured object as result data.

2. The dimension measuring device according to claim 1, wherein: The pre-trained learning model is a learning model based on deep learning.

3. The dimension measuring device according to claim 1, wherein: The instruction to output the size information of the measured object as result data includes an instruction to segment the image of interest into at least one physical object using the pre-trained learning model and mark size measurement points of the measured object.

4. The dimension measuring device according to claim 1, wherein: The instruction for acquiring the image of interest includes: Instructions for the image of interest containing the measurement object are obtained from the object image using a rule-based algorithm.

5. The dimension measuring device according to claim 4, wherein: The instruction for acquiring the image of interest includes: The instruction of obtaining the image of interest by performing image preprocessing on the object image.

6. A size measurement method based on a learning model, the size measurement method comprising the following steps: acquiring an image of interest including a measurement object whose size is to be measured from the object image; as well as The image of interest is input into a pre-trained learning model and the size information of the measured object is output as result data.

7. The dimension measuring method according to claim 6, wherein: The pre-trained learning model is a learning model based on deep learning.

8. The dimension measuring method according to claim 6, wherein: The step of outputting the size information of the measured object as result data includes the steps of segmenting the image of interest into at least one physical object using the pre-trained learning model and marking size measurement points of the measured object.

9. The dimension measuring method according to claim 6, wherein: The step of acquiring the image of interest includes the step of obtaining the image of interest including the measurement object from the object image using a rule-based algorithm.

10. The dimension measuring method according to claim 9, wherein: The step of acquiring the image of interest includes obtaining the image of interest by performing image preprocessing on the object image.

Images