KR102300500B1 - Method, device and system for processing three-dimensional cut image of product - Google Patents

Abstract

In accordance with one embodiment of the present invention, a method for processing an image of a stereoscopic picture of a product, performed by a device, comprises the steps of: obtaining a first image which is an original image of a product from a photographing device when a product worn by a mannequin or a person is photographed; generating a first input signal by encoding the first image; inputting the first input signal to a first artificial neural network, and obtaining a first output signal on the basis of a result of inputting the first artificial neural network; identifying a first category which is a type of the product on the basis of the first output signal; comparing a representative image of the first category with the first image for each part, and setting a correction value for each part of the product; generating a second image which is an image that an outline of the product is corrected by performing a pixel fluidization operation for each part in the first image on the basis of the correction value; analyzing the second image to distinguish a product area occupied by the product and a background area in the second image; and deleting the background area from the second image to generate a third image which is an image including only the product area. Therefore, a stereoscopic picture of the product can be quickly and easily provided.

Description

Product stereoscopic cut image processing method, device and system {METHOD, DEVICE AND SYSTEM FOR PROCESSING THREE-DIMENSIONAL CUT IMAGE OF PRODUCT}

The embodiments below relate to a technology for processing an image for a three-dimensional cut of a product.

Recently, as transactions through online have increased rapidly, various products are registered and sold in online shopping malls.

In an online shopping mall, various product pictures are posted on the shopping mall because the purchase is made by checking only product pictures without seeing the product directly.

However, for products worn by people, such as clothing and miscellaneous goods, photos are taken without people wearing them to reduce shooting time, so it is difficult for a buyer to understand the exact size and shape of the product.

In addition, even when a picture is taken while a person is wearing the product, since the person and the product are exposed together, there is a problem that only the product is not highlighted.

Accordingly, there is an increasing demand for providing a three-dimensional cut of a product easily and quickly so that only the product can be highlighted while displaying the three-dimensional characteristics of the product.

According to an embodiment, the first category, which is the type of the product, is identified from the photographing equipment through the first image that is the original image of the product, and the representative image of the first category and the first image are compared for each part, and each part of the product A correction value is set, a pixelation operation is performed for each part in the first image based on the correction value, a second image is generated that is an image in which the fit of the product is corrected, and the second image is analyzed to convert the product from the second image. To provide a product stereoscopic image processing method, apparatus, and system for separating the product area and the background area occupied by the product area and deleting the background area from the second image to generate a third image that is an image composed only of the product area for that purpose

Objects of the present invention are not limited to the objects mentioned above, and other objects not mentioned will be clearly understood from the description below.

According to one embodiment, in the method of processing an image for a three-dimensional cut of a product, performed by an apparatus, when a photographing of a product worn by a mannequin or a person is performed, a first image that is an original image of the product from a photographing device acquiring an image; generating a first input signal by encoding the first image; inputting the first input signal to a first artificial neural network, and obtaining a first output signal based on a result of the input of the first artificial neural network; identifying a first category that is a type of the product based on the first output signal; comparing the representative image of the first category with the first image for each part, and setting a correction value for each part of the product; generating a second image in which the fit of the product is corrected by performing a pixel fluidization operation for each part in the first image based on the correction value; analyzing the second image to distinguish a product area and a background area occupied by the product in the second image; and deleting the background area from the second image to generate a third image that is an image composed of only the product area.

The product three-dimensional cut image processing method may include: recognizing an outline of the product in the third image, and dividing the outline of the product into sections preset in the first category; classifying each of the plurality of lines divided for each section in the outline of the product into a straight line and a curved line; Recognizes each of a plurality of points located at a point where each of the plurality of lines is connected, and a first type in which a straight line and a straight line are connected to each of the plurality of points, a second type in which a straight line and a curved line are connected, and a curve and a straight line are connected to each of the plurality of points Classifying into a third type that becomes a curve and a fourth type in which the curve is connected; and registering information on the order, coordinates, and type of each of the plurality of points as characteristic information of the product.

The step of classifying each of the plurality of lines into a straight line and a curved line may include: recognizing a first point and a second point located at both ends of a first line that is any one of the plurality of lines; recognizing a third point located in any part of the middle other than both ends of the first line; generating a second line connecting the first point and the third point with a straight line and a third line connecting the second point and the third point with a straight line; checking whether a fourth line combining the second line and the third line matches the first line; and when it is confirmed that the fourth line and the first line coincide, the first line is classified as a straight line, and when it is confirmed that the fourth line and the first line do not match, the first line is classified as a curve may include the step of

The product three-dimensional cut image processing method may include: calculating a first brightness value that is an average brightness value of pixels in the product area; calculating a second brightness value that is an average brightness value of pixels in the background area; calculating a first difference value that is a difference between the first brightness value and the second brightness value; determining that lighting control of the photographing equipment is necessary when it is determined that the first difference value is out of a preset first reference range; calculating a second difference value that is a difference between the first difference value and the minimum value of the first reference range when it is determined that the first difference value is smaller than the minimum value of the first reference range; changing the brightness setting level for the lighting of the photographing equipment to a higher level than the currently set level based on the second difference value; calculating a third difference value that is a difference between the first difference value and the maximum value of the first reference range when it is determined that the first difference value is greater than the maximum value of the first reference range; and changing the brightness setting level for the lighting of the photographing equipment to a level lower than the currently set level based on the third difference value.

The product three-dimensional cut image processing method includes: checking a target temperature and a target humidity set in each of a plurality of photographing equipment disposed in a photographing room; By controlling the operation of each of the plurality of thermo-hygrostat connected to each of the plurality of photographing equipment, the temperature measured by each of the plurality of photographing equipment is adjusted to maintain the target temperature, and the humidity measured by each of the plurality of photographing equipment is adjusted adjusting to maintain the target humidity; checking the number of adjustments adjusted to maintain the target temperature and the target humidity for a preset period for each of the plurality of photographing devices; Classifying the first photographing equipment in which the number of adjustments is confirmed to be within the preset second reference range is classified as a normal state without abnormality, and the second photographing equipment in which the number of adjustments is determined to be out of the second reference range is abnormal. classifying it as an abnormal state; when the second imaging device is classified as an abnormal state, controlling to transmit a notification message regarding the abnormal state of the second imaging device to a manager terminal; When it is determined that the number of adjustments of the second photographing equipment is smaller than the minimum value of the second reference range, the second photographing equipment is classified as an idle state from an abnormal state, and the number of adjustments of the second photographing equipment is determined by the second reference classifying the second imaging device from an abnormal state to an overload state when it is confirmed that the value is greater than the maximum value of the range; When the second photographing equipment is classified as an idle state, a notification message regarding the idle state of the second photographing equipment is controlled to be transmitted to the manager terminal, and when the second photographing equipment is classified as an overload state, the second photographing controlling a notification message for an overload state of the equipment to be transmitted to the manager terminal; If it is confirmed that the idle state of the second photographing equipment continues for more than a preset reference time, the second photographing equipment is classified as a replacement target equipment, and the overload state of the second photographing equipment continues for more than the reference time classifying the second photographing equipment as repair target equipment when it is confirmed that the and when the second photographing equipment is classified as a replacement target device, a notification message informing that a replacement of a seat for the second photographing equipment is required is transmitted to the manager terminal, and the second photographing equipment is the equipment to be repaired The method may further include controlling to transmit a notification message informing that repair of the second photographing equipment is required to be transmitted to the manager terminal.

According to an embodiment, the first category, which is the type of the product, is identified from the photographing equipment through the first image that is the original image of the product, and the representative image of the first category and the first image are compared for each part, and each part of the product A correction value is set, a pixelation operation is performed for each part in the first image based on the correction value, a second image is generated that is an image in which the fit of the product is corrected, and the second image is analyzed to convert the product from the second image. By separating the product area and the background area occupied by this and deleting the background area from the second image, the third image is created, which is an image composed of only the product area, so that only the product can stand out while showing the three-dimensional characteristics of the product. It has the effect of providing a three-dimensional cut of the product quickly and easily.

On the other hand, effects according to the embodiments are not limited to those mentioned above, and other effects not mentioned will be clearly understood by those of ordinary skill in the art from the following description.

1 is a flowchart for explaining a process of processing an image for a three-dimensional product cut according to an embodiment.
2 is a flowchart illustrating a process of analyzing the appearance characteristics of a product according to an exemplary embodiment.
3 is a diagram for explaining a process of recognizing a plurality of lines and a plurality of points through an outline of a product according to an exemplary embodiment.
4 is a flowchart illustrating a process of classifying each of a plurality of lines into a straight line and a curved line according to an exemplary embodiment.
5 is a flowchart illustrating a process of classifying a first line into a straight line according to an exemplary embodiment.
6 is a flowchart illustrating a process of classifying a first line into a curve according to an exemplary embodiment.
7 is a flowchart illustrating a process of controlling lighting of a photographing device according to an exemplary embodiment.
8 is a flowchart illustrating a process of classifying states of photographing equipment according to an exemplary embodiment.
9 is a diagram for explaining a first artificial neural network according to an embodiment.
10 is a diagram for explaining a method of learning a first artificial neural network according to an embodiment.
11 is an exemplary diagram of a configuration of an apparatus according to an embodiment.

Hereinafter, embodiments will be described in detail with reference to the accompanying drawings. However, since various changes may be made to the embodiments, the scope of the patent application is not limited or limited by these embodiments. It should be understood that all modifications, equivalents and substitutes for the embodiments are included in the scope of the rights.

Specific structural or functional descriptions of the embodiments are disclosed for purposes of illustration only, and may be changed and implemented in various forms. Accordingly, the embodiments are not limited to a specific disclosure form, and the scope of the present specification includes changes, equivalents, or substitutes included in the technical spirit.

Although terms such as first or second may be used to describe various components, these terms should be interpreted only for the purpose of distinguishing one component from another. For example, a first component may be termed a second component, and similarly, a second component may also be termed a first component.

When a component is referred to as being “connected to” another component, it may be directly connected or connected to the other component, but it should be understood that another component may exist in between.

The terms used in the examples are used for description purposes only, and should not be construed as limiting. The singular expression includes the plural expression unless the context clearly dictates otherwise. In the present specification, terms such as “comprise” or “have” are intended to designate that a feature, number, step, operation, component, part, or combination thereof described in the specification exists, but one or more other features It should be understood that this does not preclude the existence or addition of numbers, steps, operations, components, parts, or combinations thereof.

Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which the embodiment belongs. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with the meaning in the context of the related art, and should not be interpreted in an ideal or excessively formal meaning unless explicitly defined in the present application. does not

In addition, in the description with reference to the accompanying drawings, the same components are given the same reference numerals regardless of the reference numerals, and the overlapping description thereof will be omitted. In the description of the embodiment, if it is determined that a detailed description of a related known technology may unnecessarily obscure the gist of the embodiment, the detailed description thereof will be omitted.

The embodiments may be implemented in various types of products, such as personal computers, laptop computers, tablet computers, smart phones, televisions, smart home appliances, intelligent cars, kiosks, wearable devices, and the like.

1 is a flowchart for explaining a process of processing an image for a three-dimensional product cut according to an embodiment.

According to an embodiment, the apparatus is an apparatus for performing a method of processing an image for a three-dimensional cut of a product, and may be implemented as, for example, a hardware module, a software module, or a combination thereof.

According to an embodiment, the device may be implemented as a terminal, and the device may be implemented as a server. The server may communicate with the terminal, retrieve necessary information from the database, employ it in the judgment operation, and update the database based on the result of the judgment. The operation performed by the apparatus may be implemented by either one of the server and the terminal alone, or may be implemented by the interaction of the server and the terminal. The server and the terminal may include an electronic device.

According to an embodiment, the electronic device may be a smart home appliance having a communication function. Smart home appliances include, for example, a television, a digital video disk (DVD) player, an audio device, a refrigerator, an air conditioner, a vacuum cleaner, an oven, a microwave oven, a washing machine, an air purifier, a set-top box, and a TV. It may include at least one of a box (eg, Samsung HomeSync™, Apple TV™, or Google TV™), game consoles, an electronic dictionary, an electronic key, a camcorder, or an electronic picture frame.

According to an embodiment, the electronic device includes various medical devices (eg, magnetic resonance angiography (MRA), magnetic resonance imaging (MRI), computed tomography (CT), imagers, ultrasound machines, etc.), navigation devices, and GPS receivers ( Global positioning system receiver), EDR (event data recorder), FDR (flight data recorder), automotive infotainment device, marine electronic equipment (e.g. marine navigation system and gyro compass, etc.), avionics, security It may include at least one of a device, a head unit for a vehicle, an industrial or household robot, an automatic teller's machine (ATM) of a financial institution, or a point of sales (POS) of a store.

According to one embodiment, the electronic device is a piece of furniture or a building/structure including a communication function, an electronic board, an electronic signature receiving device, a projector, or various measurement devices. It may include at least one of devices (eg, water, electricity, gas, or radio wave measuring device, etc.). The electronic device according to an embodiment may be a combination of one or more of the various devices described above. Also, the electronic device according to an embodiment may be a flexible device. Also, it is apparent to those skilled in the art that the electronic device according to an embodiment is not limited to the above-described devices. The term user used in various embodiments may refer to a person who uses an electronic device or a device (eg, an artificial intelligence electronic device) using the electronic device.

An electronic device according to an embodiment includes a processor, a memory, a user interface, and a communication interface, and may be connected to another electronic device through a network. The communication interface may transmit/receive data to and from another electronic device within a predetermined distance through a wired, wireless network, or wired serial communication. The network enables wired and wireless communication between the electronic device and various entities according to an embodiment. The electronic device may communicate with various entities over a network, and the network may use standard communication technologies and/or protocols. In this case, the network includes, but is not limited to, the Internet, a local area network (LAN), a wireless local area network (Wireless LAN), a wide area network (WAN), a personal area network (PAN), and the like, A person of ordinary skill in the art of communication technology can recognize that there may be other types of networks capable of transmitting and receiving information.

The portable terminal according to an embodiment may be an electronic device including a communication function. For example, the electronic device includes a smart phone, a tablet personal computer (PC), a mobile phone, a video phone, an e-book reader, a desktop personal computer (PC), and a laptop. PC (laptop personal computer), netbook computer (netbook computer), PDA (personal digital assistant), PMP (portable multimedia player), MP3 player, mobile medical device, camera, or wearable device (such as: at least one of a head-mounted-device (HMD) such as electronic glasses, electronic apparel, electronic bracelets, electronic necklaces, electronic accessories, electronic tattoos, smart cars, or smartwatches. can

According to an embodiment, the device may be connected to the photographing equipment through wired/wireless and control the operation of the photographing equipment. Here, the photographing equipment may be photographing equipment for photographing a three-dimensional cut. A three-dimensional cut is an image showing the three-dimensional characteristics of a mannequin or a product worn by a person.

When photographing equipment takes a picture of a product, the device acquires an image of the product picture, recognizes the border of the product on the image, removes the background area other than the product by itself, and can freely extract the file format of the product image. Therefore, you can take product photos without the need for a separate photographer.

The device may provide functions such as background removal, left-right symmetry, wrinkle removal, and mannequin removal through photoshop processing of the photographed product photo. At this time, the device may process the image by capturing the fit worn by the product, and may provide a service to synthesize by working only half of the product photo or to synthesize the material of the product.

In determining the fit of the product, the device may set a standard through the image of the product worn by the mannequin, and may process the process of fixing the basic fit of the product in a different way for each product category.

According to an embodiment, the photographing equipment checks the characteristics of the product when taking a three-dimensional cut of the product under the control of the device, automatically sets the lighting of the photographing equipment, and shoots the product from various angles to obtain a three-dimensional image can create In this case, the product may include various products such as clothes, shoes, and hats.

Referring to FIG. 1 , first, in step S101, when a photographing of a product worn by a mannequin or a person is performed through the photographing equipment, the device may acquire a first image that is an original image of the product from the photographing equipment.

The device may encode the first image to generate a first input signal.

Specifically, the device may generate the first input signal by encoding the pixels of the first image with color information. The color information may include, but is not limited to, RGB color information, brightness information, and saturation information. The device may convert the color information into a numerical value, and may encode the first image in the form of a data sheet including the value.

The device may input the first input signal to the pre-trained first artificial neural network in the device.

The first artificial neural network according to an embodiment is composed of a feature extraction neural network and a classification neural network, and the feature extraction neural network sequentially stacks a convolutional layer and a pooling layer on an input signal. The convolution layer includes a convolution operation, a convolution filter, and an activation function. The calculation of the convolution filter is adjusted according to the matrix size of the target input, but a 9X9 matrix is generally used. The activation function generally uses, but is not limited to, a ReLU function, a sigmoid function, and a tanh function. The pooling layer is a layer that reduces the size of the input matrix, and uses a method of extracting representative values by tying pixels in a specific area. In general, the average value or the maximum value is often used for the calculation of the pooling layer, but is not limited thereto. The operation is performed using a square matrix, usually a 9x9 matrix. The convolutional layer and the pooling layer are repeated alternately until the corresponding input becomes small enough while maintaining the difference.

According to an embodiment, the classification neural network has a hidden layer and an output layer. Classification of the first artificial neural network for object candidate detection The neural network consists of five or less hidden layers, and may include a total of 50 or less hidden layer nodes. The activation function of the hidden layer uses a ReLU function, a sigmoid function, and a tanh function, but is not limited thereto. There is a total of one output layer node of the classification neural network, and the output value for the type of product can be output to the output layer node. A detailed description of the first artificial neural network will be described later with reference to FIG. 9 .

The device may obtain a first output signal based on a result of the input of the first artificial neural network.

In step S102 , the device may identify a first category, which is a type of product, based on the first output signal.

For example, as a result of checking the output value of the first output signal, when the output value is 1, the device identifies the type of product as a category on clothes, and when the output value is 2, identifies the type of product as a category under clothes, When the output value is 3, the type of product can be identified as a shoe category, and when the output value is 4, the type of product can be identified as a hat category.

In step S103, the device may obtain the representative image of the first category from the database. To this end, representative images registered for each category may be stored in the database of the device.

In step S104 , the device may compare the representative image of the first category and the first image for each part, and set a correction value for each part of the product.

Specifically, when the type of product is identified as the first category through the first image, the device matches the representative image of the first category with the file extension format of the first image, and also matches the scale and resolution, so that the size and A preprocessing process can be performed to adjust the shape to some extent. The device performs image comparison for each part of the product, and as a result of the image comparison, it is possible to identify parts with differences in size and shape, and a correction value that matches the product in the first image with the product in the representative image of the first category can be set for each part. Here, the correction value is a setting value applied when performing the pixel fluidization operation, and may include setting values such as expansion, reduction, and movement of a specific part.

That is, the device adjusts the size and shape of the product in the first image to match the size and shape of the product in the representative image of the first category, so that the fit of the product for each type of product can be set through a unified standard. have.

In step S105 , the device may generate a second image that is an image in which the fit of the product is corrected by performing a pixel fluidization operation for each part in the first image based on the correction value.

For example, when the type of product is clothes top, correction values may be set for each part such as sleeve, arm, torso, neck, etc., and as a result of comparing the first image with the representative image of the first category, When there is a large difference for the sleeve region, the correction value for the sleeve region may be set to a large value, and when there is no difference for the neck region, the correction value for the neck region may be set to a small value. The device performs a pixelation operation on the sleeve region in the first image through the correction value for the sleeve region, performs image correction through a large value for the sleeve region, and performs image correction through the correction value for the neck region to the first image By performing a pixelation operation on the neck region, image correction can be performed with a small value for the neck region.

That is, the device corrects the size and shape of the product in the first image for each part so as to match the size and shape of the product in the representative image of the first category, and is different from the product in the representative image of the first category As the number increases, image correction is performed through a larger setting value, and image correction can be performed through a smaller setting value as it is similar to a product in the representative image of the first category.

The device may generate a second image that is an image in which the fit of the product is corrected by calibrating each part of the product in the first image, and the product in the second image is the same size as the product in the representative image of the first category And the shape may be in a corrected state with the same fit.

In step S106 , the device may analyze the second image to distinguish a product area occupied by the product and a background area in the second image.

Specifically, the device may recognize the outline of the product in the second image, classify an area inside the outline of the product as a product area, and classify an area outside the outline of the product as a background area.

In step S107 , the device may delete the background area from the second image to generate a third image that is an image composed of only the product area.

According to an exemplary embodiment, work efficiency may be increased when photographing a product of a different color by using the third image, which is a work result that has been previously used. That is, when creating a frame (frame) of a product through the third image and adding an image for a product of a different color using the frame, an image with a different color is added by substituting the third image, which is the existing result. By acquiring as , consistent images can be expressed in the case of products with the same shape but different colors.

Specifically, the device acquires a 1-1 image that is an original image of the first product, and generates a 1-2 image in which the first product is an image in which the fit is corrected through the 1-1 image, and the first- Through the second image, it is possible to generate images 1-3 consisting of only the area of the first product. In this case, the 1-3 images may be used as a frame indicating the shape of the first product.

Thereafter, the device may acquire a 2-1 image, which is an original image of a second product that has the same shape as the first product but has a different color, and is an image in which the fit of the second product is corrected through the 2-1 image. Even without performing the operation of generating the 2-2 image, the 2-3 image including only the area of the second product may be generated through the 1-3 image. In this case, the 2-3 image is the color of the first product changed to the color of the second product in the 1-3 image, and even if the work of correcting the fit for the second product is not performed, the first An image of the second product may be easily obtained by using the image of the product.

Accordingly, there is an effect of increasing the speed of the device in generating a work result for each product image by photographing several products. That is, the work of correcting the fit of the product is performed only when the work result for the first product is created, and when the work result for the second product, the third product, etc. is generated, the work result for the first product is used Therefore, it is possible to quickly generate work results, so that a large number of work results can be obtained quickly.

According to an embodiment, after generating a third image, which is an image consisting of only the product area, in order to accurately match the color of the product identified in the third image and the color of the real product, the color of the real is fully corrected and expressed. processing may be performed.

Specifically, in acquiring the 1-1 image, which is the original image of the first product, the device may acquire a plurality of images obtained by photographing the first product from various angles. For example, a 1-1-1 image that is an image of the first product photographed from a first angle, a 1-2-1 image that is an image of the first product photographed from a second angle, and the like may be acquired. At this time, if the first angle is set as the main angle, the device generates a 1-2 image, which is an image in which the fit of the first product is corrected, through the 1-1-1 image, and displays the 1-2 image. Through this, it is possible to generate the 1-3 image composed of only the region of the first product.

Thereafter, when the device compares the color value of the 1-3 image and the color value of the 1-1-2 image and determines that the difference value is out of the reference range, the color value of the 1-3 image and the first The average value of the color values of the -1-2 image may be calculated, and the color of the image 1-3 may be changed through the calculated average value.

That is, by performing a process of comparing the colors of images photographed from various angles and matching them to an average value, it is possible to process the color so that the same color as the actual color of the product is applied as much as possible.

2 is a flowchart illustrating a process of analyzing the appearance characteristics of a product according to an exemplary embodiment.

Referring to FIG. 2 , first, in step S201 , the device may recognize the outline of the product in the third image. Since the third image consists of only the product area, the edge of the image may be recognized as the outline of the product.

In step S202, the device may divide the outline of the product into sections preset in the first category.

For example, the device has an interval set for each part according to the type of product, so the outline of the product can be divided into sections according to the preset interval, and a part that changes by more than a specific angle is determined as the start of a new section. The outline of the product can be divided into sections.

In step S203, the device may recognize each of the plurality of lines divided by section in the outline of the product, and in step S204, the device may recognize each of the plurality of points located at a point where each of the plurality of lines is connected. A detailed description related thereto will be described later with reference to FIG. 3 .

In step S205, the device may classify each of the plurality of lines recognized in step S203 by type. Line types may be classified into straight lines and curved lines, and the device may classify each of the plurality of lines into straight lines and curved lines. A detailed description related thereto will be described later with reference to FIG. 4 .

In step S206, the device may classify each of the plurality of points recognized in step S204 by type. The type of point may be divided into a first type in which a straight line and a straight line are connected, a second type in which a straight line and a curve are connected, a third type in which a curve and a straight line are connected, and a fourth type in which a curve and a curve are connected, may classify each of the plurality of points into a first type, a second type, a third type, and a fourth type.

In step S207 , the device may extract and register information on the sequence number, coordinates, and type of each of the plurality of points as product feature information.

For example, when the first point, the second point, and the third point are recognized through the third image, the sequence number connecting the first point, the second point, and the third point, and the first point, the second point, and the second point Coordinates where each of the three points are located, and information about the type in which each of the first point, the second point, and the third point are classified are extracted as feature information of the product in the third image, and the extracted feature information is combined with the third image They can be linked and stored in the database.

3 is a view for explaining a process of recognizing a plurality of lines and a plurality of points through an outline of a product according to an exemplary embodiment.

First, as shown in (a) of FIG. 3 , the device may recognize the outline 300 of the product in the third image. In this case, in the third image, the outline 300 of the product is constituted as an edge, and the background area is deleted.

Thereafter, the device may divide the outline 300 of the product into sections preset in the first category.

Specifically, as shown in (b) of FIG. 3 , the device draws the outline 300 of the product for each section preset in the clothing category, the first line 311 , the second line 312 , and the third line. (313), the fourth line (314), the fifth line (315), the sixth line (316), the seventh line (317), the eighth line (318), the ninth line (319), the tenth line ( 320 ), an eleventh line 321 , and a twelfth line 322 . The outline 300 of the product may be divided into more or fewer sections depending on the type of product.

The device identifies each of the plurality of lines, the second line 312 , the third line 313 , the fourth line 314 , the fifth line 315 , the sixth line 316 , and the seventh line 317 . ), the 8th line 318 , the 9th line 319 , and the 10th line 320 are classified as straight lines, and the 1st line 311 , the 11th line 321 and the 12th line 322 are curved. can be classified as

The device may recognize each of a plurality of points located at a point where each of the plurality of lines is connected.

Specifically, as shown in (c) of FIG. 3 , the device includes a first point 331 located at a point where the twelfth line 322 and the first line 311 are connected, the first line 311 and A second point 332 positioned at a point where the second line 312 is connected, a third point 333 positioned at a point where the second line 312 and the third line 313 are connected, and a third line 313 ) and the fourth point 334 positioned at the point where the fourth line 314 is connected, the fifth point 335 positioned at the point where the fourth line 314 and the fifth line 315 are connected, the fifth line The sixth point 336 located at the point where 315 and the sixth line 316 are connected, the seventh point 337 located at the point where the sixth line 316 and the seventh line 317 are connected, the th The eighth point 338 located at the point where the seventh line 317 and the eighth line 318 are connected, and the ninth point 339 located at the point where the eighth line 318 and the ninth line 319 are connected. , the tenth point 340 located at the point where the ninth line 319 and the tenth line 320 are connected, and the eleventh point located at the point where the tenth line 320 and the eleventh line 321 are connected ( 341 ) and a twelfth point 342 located at a point where the eleventh line 321 and the twelfth line 322 are connected may be recognized. In this case, the positions of the first point 331 serving as the starting point and the twelfth point 342 serving as the end point may be set differently for each category according to the type of product.

The device may classify the type of each of the plurality of points by identifying the type of the line connected to each of the plurality of points.

Specifically, the third point 333 , the fourth point 334 , the fifth point 335 , the sixth point 336 , the seventh point 337 , the eighth point 338 , and the ninth point 339 . ) and the tenth point 340 are classified as a first type because a straight line and a straight line are connected, and the eleventh point 341 is classified as a second type because a straight line and a curved line are connected, and a second point 332 ). The curve and the straight line are connected to be classified as a third type, and the first point 331 and the twelfth point 342 may be classified as a fourth type because the curve and the curve are connected.

Thereafter, the device determines a first point 331 , a second point 332 , a third point 333 , a fourth point 334 , a fifth point 335 , a sixth point 336 , a seventh point ( 337), the eighth point 338, the ninth point 339 and the tenth point 340, the eleventh point 341 and the twelfth point 342, respectively, through information on the order, coordinates and types of them, The product feature information may be extracted, and the extracted product feature information may be registered as the product feature information in the third image and stored in a database.

4 is a flowchart illustrating a process of classifying each of a plurality of lines into a straight line and a curved line according to an exemplary embodiment.

Referring to FIG. 4 , first, in step S401 , the device may recognize a first line that is any one of a plurality of lines. At this time, the device divides the outline of the product recognized in the third image into sections, and a plurality of lines are formed.

In step S402, the device may recognize the first point and the second point located at both ends of the first line.

In step S403 , the device may recognize the third point located at any part in the middle instead of at both ends of the first line. The position of the third point may be set differently depending on the embodiment.

In step S404 , the device may generate a second line connecting the first point and the third point with a straight line, and may generate a third line connecting the second point and the third point with a straight line.

In step S405 , the device may recognize a fourth line combining the second line and the third line.

In step S406 , the device may check whether the first line and the fourth line match.

If it is determined in step S406 that the first line and the fourth line coincide with each other, in step S407 , the device may classify the first line as a straight line.

If it is determined in step S406 that the first line and the fourth line do not match, in step S408, the device may classify the first line as a curve.

5 is a flowchart illustrating a process of classifying a first line into a straight line according to an exemplary embodiment.

First, as shown in (a) of FIG. 5 , the device may recognize a first line that is any one of a plurality of lines generated by dividing an outline of a product into sections.

As shown in FIG. 5B , the device may recognize the first point and the second point positioned at both ends of the first line, and recognize the third point positioned in the middle of the first line. The position of the third point may be set differently depending on the embodiment.

As shown in FIG. 5C , the device generates a second line connecting the first and third points with a straight line, and a third line connecting the third and second points with a straight line. can create

As shown in FIG. 5D , the device may recognize a fourth line combining the second line and the third line.

Thereafter, the device compares the first line shown in FIG. 5A and the fourth line shown in FIG. 5D to confirm that the first line and the fourth line coincide, and the first line and the second line Since the four lines coincide, the first line can be classified as a straight line.

6 is a flowchart illustrating a process of classifying a first line into a curve according to an exemplary embodiment.

First, as shown in (a) of FIG. 6 , the device may recognize a first line that is any one of a plurality of lines generated by dividing an outline of a product into sections.

As shown in FIG. 6B , the device may recognize the first point and the second point positioned at both ends of the first line, and recognize the third point positioned in the middle of the first line. The position of the third point may be set differently depending on the embodiment.

As shown in FIG. 6C , the device generates a second line connecting the first and third points with a straight line, and a third line connecting the third and second points with a straight line. can create

As shown in (d) of FIG. 6 , the device may recognize a fourth line combining the second line and the third line.

Thereafter, the device compares the first line shown in (a) of FIG. 6 and the fourth line shown in FIG. 6 (d) to confirm that the first line and the fourth line do not match, and the first line and the fourth line are not identical. Since the fourth line does not coincide, the first line can be classified as a curve.

7 is a flowchart illustrating a process of controlling lighting of a photographing device according to an exemplary embodiment.

Referring to FIG. 7 , first, in step S701 , the device analyzes the second image to distinguish a product area occupied by the product in the second image and a background area that is other areas.

In step S702 , the device may calculate a first brightness value that is an average brightness value of pixels in the product area. In this case, the device may calculate the first brightness value by checking brightness information of pixels in the product area.

In operation S703 , the device may calculate a second brightness value that is an average brightness value of pixels in the background area. In this case, the device may calculate the second brightness value by checking brightness information of pixels in the background area.

In step S704 , the device may calculate a first difference value that is a difference between the first brightness value and the second brightness value.

In step S705 , the device may determine whether the first difference value is within the first reference range. Here, the first reference range may be set differently according to embodiments.

If it is determined in step S705 that the first difference value is within the first reference range, the device determines that the lighting control of the photographing equipment is not necessary, and can process to maintain the brightness setting level set in the lighting of the photographing equipment. have.

For example, when the first reference range is set to 4 or more and 6 or less, when the first difference value is 5, the device may determine that lighting control of the photographing equipment is not required.

If it is determined in step S705 that the first difference value is out of the first reference range, in step S706 , the device may determine whether the first difference value is smaller than the minimum value of the first reference range.

If it is determined in step S706 that the first difference value is smaller than the minimum value of the first reference range, in step S707 , the device may calculate a second difference value that is a difference between the first difference value and the minimum value of the first reference range.

In step S708 , the device may change the brightness setting level for the lighting of the photographing equipment to a higher level than the currently set level based on the second difference value.

For example, if the brightness setting level for the lighting of the photographing equipment is currently set to level 5 and the second difference value is 1, the brightness setting level may be changed from level 5 to level 6, and the second difference value is In the case of 2, the brightness setting level can be changed from level 5 to level 7.

That is, if the difference in brightness between pixels in the product area and pixels in the background area is less than the reference range, change the brightness setting for the lighting of the shooting equipment to a higher level than the currently set level. It can be controlled to proceed with product photography through bright lighting.

If it is determined in step S706 that the first difference value is not smaller than the minimum value of the first reference range, since the first difference value is out of the first reference range, it is determined that the first difference value is greater than the maximum value of the first reference range. may be confirmed, and in step S709 , the device may calculate a third difference value that is a difference between the first difference value and the maximum value of the first reference range.

In step S710 , the device may change the brightness setting level for the lighting of the photographing equipment to a level lower than the currently set level based on the third difference value.

For example, if the brightness setting level for the lighting of the photographing equipment is currently set to level 5 and the third difference value is 1, the brightness setting level may be changed from level 5 to level 4, and the third difference value is In the case of 2, the brightness setting level can be changed from level 5 to level 3.

That is, if the difference in brightness between the pixels in the product area and the pixels in the background area exceeds the reference range, change the brightness setting for the lighting of the shooting equipment to a lower level than the currently set level. It can be controlled to proceed with product photography through dim lighting.

8 is a flowchart illustrating a process of classifying states of photographing equipment according to an exemplary embodiment.

According to an embodiment, the system may further include a sensor group and a thermo-hygrostat in addition to the plurality of imaging equipment and devices.

The system may include a plurality of photographing equipment disposed in the photographing room, for example, when the first photographing equipment, the second photographing equipment, and the third photographing equipment are disposed in the photographing room, in response to the first photographing equipment The first sensor group and the first thermo-hygrostat are further included, the second sensor group and the second thermo-hygrostat are further included to correspond to the second imaging device, and the third sensor group and the third thermo-hygrostat to correspond to the third imaging device. A humidifier may be further included.

Each of the plurality of photographing equipment may include a sensor group for measuring the temperature and humidity of the internal air. For example, the first imaging device includes a first sensor group for measuring the temperature and humidity of the internal air of the first imaging device, and the second imaging device is configured to measure the temperature and humidity of the internal air of the second imaging device A second sensor group may be included.

Each of the plurality of sensor groups may include a temperature sensor for measuring temperature and a humidity sensor for measuring humidity.

A plurality of thermo-hygrostat may be connected to each of the plurality of imaging devices, for example, a first thermo-hygrostat may be connected to the first imaging device, and a second thermo-hygrostat may be connected to the second imaging device.

The first thermo-hygrostat may heat or cool the air to maintain the target temperature set in the first imaging device, and humidify or dehumidify the air to maintain the target humidity set in the first imaging device.

Similarly, the second thermostat may heat or cool the air to maintain the target temperature set in the second imaging device, and humidify or dehumidify the air to maintain the target humidity set in the second imaging device.

According to an embodiment, the target temperature set in the first imaging device and the target temperature set in the second imaging device may be automatically set to be the same, and each target temperature may be set differently according to a user setting.

In addition, the target humidity set in the first imaging device and the target humidity set in the second imaging device may be automatically set to be the same, and each target humidity may be set differently according to a user setting.

The device may control operations of the first imaging device, the second imaging device, the first sensor group, the second sensor group, the first thermo-hygrostat, and the second thermo-hygrostat included in the system to be normally performed.

The device controls the operation of each of the first and second thermo-hygrostats to adjust the temperatures measured by each of the first and second imaging devices to maintain target temperatures, and the first and second imaging devices Humidity measured in each can be adjusted to maintain the target humidity. In this case, the first sensor group may measure the temperature and humidity of the first imaging device, and the second sensor group may measure the temperature and humidity of the second imaging device.

Meanwhile, according to an embodiment, the device checks the temperature measured by the first sensor group, controls the operation of the first thermo-hygrostat connected to the first imaging device to maintain the target temperature set in the first imaging device, and By checking the humidity measured in the first sensor group, the operation of the first thermo-hygrostat may be controlled to maintain the target humidity set in the first imaging device.

In addition, the device checks the temperature measured by the second sensor group, controls the operation of the second thermo-hygrostat connected to the second imaging device to maintain the target temperature set in the second imaging device, and By checking the humidity, the operation of the second thermo-hygrostat may be controlled to maintain the target humidity set in the second imaging device.

According to an embodiment, the device calculates a first temperature difference value that is the difference between the temperature measured by the first sensor group and the target temperature set in the first imaging device, and when it is confirmed that the first temperature difference value is greater than the reference value, The first thermo-hygrostat connected to the first imaging device may be controlled to operate.

For example, when the temperature measured by the first sensor group is confirmed to be 30 ° C, and the target temperature set in the first imaging device is confirmed to be 25 ° C, the device may calculate 5 ° C as the first temperature difference value, , when the reference value is set to 3° C., it is possible to control the first thermo-hygrostat to operate because a temperature change is required.

In addition, the device calculates a second temperature difference value that is a difference between the temperature measured by the second sensor group and the target temperature set in the second imaging device, and when it is confirmed that the second temperature difference value is greater than the reference value, the second imaging device It is possible to control the second thermo-hygrostat connected to the operation.

According to an embodiment, the device may check the target temperature and target humidity set in each of a plurality of imaging equipment disposed in the imaging room.

The device controls the operation of each of the plurality of thermo-hygrostats connected to each of the plurality of photographing equipment to adjust the temperature measured by each of the plurality of photographing equipment to maintain the target temperature, and the humidity measured by each of the plurality of photographing equipment is the target humidity can be adjusted to maintain

In step S801, the device may check the number of adjustments adjusted to maintain the target temperature and target humidity for a preset period for each of the plurality of photographing equipment.

For example, when a temperature change is required for the first imaging device, the device may control the first thermo-hygrostat to operate, so that the first imaging device maintains the target temperature, and the first imaging device maintains the target temperature for a week. It can be confirmed by calculating the number of adjustments adjusted to maintain .

In step S802, the device may check whether the number of adjustments is within the second reference range.

If it is determined in step S802 that the number of adjustments is within the second reference range, in step S803, the device may classify the photographing equipment whose number of adjustments is found to be within the second reference range as a normal state with no abnormality.

For example, when the second reference range is set to 10 times or more and 20 times or less, the device may classify the first imaging device as a normal state when the number of adjustments of the first imaging device is 15 times.

If it is determined in step S802 that the number of adjustments is out of the second reference range, in step S804, the device may classify the photographing equipment in which the number of adjustments is out of the second reference range as an abnormal state.

For example, when the second reference range is set to 10 times or more and 20 times or less, the device may classify the second imaging device as an abnormal state when the number of adjustments of the second imaging device is 5 times, When the number of adjustments of the second imaging device is confirmed to be 25, the second imaging device may be classified as an abnormal state.

In step S805 , the device may determine whether the number of adjustments of the second imaging device classified as abnormal is less than the minimum value of the second reference range.

If it is determined in step S805 that the number of adjustments of the second imaging device is less than the minimum value of the second reference range, in step S806, the device may classify the second imaging device from the abnormal state to the idle state.

For example, when the second reference range is set to 10 times or more and 20 times or less, the device confirms that the number of adjustments of the second imaging equipment is 5 times, and the number of adjustments is higher than the minimum value of the second reference range of 10 times. If it is confirmed that it is small, the second imaging device may be classified as an idle state.

If it is determined in step S805 that the number of adjustments of the second photographing equipment is not less than the minimum value of the second reference range, since the number of adjustments of the second photographing equipment is out of the second reference range, the device determines the number of adjustments of the second photographing equipment is greater than the maximum value of the second reference range, and if it is determined that the number of adjustments of the second imaging device is greater than the maximum value of the second reference range, in step S807, the device removes the second imaging device from the abnormal state. It can be classified as an overload condition.

For example, if the second reference range is set to 10 times or more and 20 times or less, the device determines that the number of adjustments of the second imaging equipment is 25 times, and the number of adjustments is greater than the maximum value of the second reference range of 20 times. When it is confirmed that is large, the second imaging device may be classified as an overload state.

If the second imaging device is classified as the valid state, in step S808 , the device may determine whether the idle state of the second imaging device continues for a reference time or longer. Here, the reference time may be set differently according to embodiments.

If it is confirmed in step S808 that the idle state of the second imaging device continues for more than the reference time, in step S810 , the device may classify the second imaging device as a replacement target device.

For example, if it is confirmed that the idle state of the second imaging device has been maintained for one month or more, the device may classify the second imaging device as a replacement target device.

If the second photographing equipment is classified as an overload state, in step S809 , the device may determine whether the overload state of the second photographing equipment continues for more than a reference time.

If it is confirmed in step S809 that the overload state of the second imaging device continues for more than the reference time, in step S811 , the device may classify the second imaging device as the equipment to be repaired.

For example, if it is confirmed that the overload state of the second imaging device has continued for one month or more, the device may classify the second imaging device as a repair target device.

In step S812 , the device may control so that status notification messages for the first and second photographing equipment are transmitted to the manager terminal.

When the first imaging device is classified as a normal state in step S803, the device may control to transmit a status notification message indicating that the first imaging device is in a normal state with no abnormality to be transmitted to the manager terminal.

If the second imaging device is classified as an abnormal state in step S804, the device may control to transmit a status notification message indicating that the second imaging device is in an abnormal state to be transmitted to the manager terminal.

If it is confirmed in step S806 that the second imaging device is in the idle state, and it is confirmed that the idle state does not last longer than the reference time in step S808, the device provides a status notification message indicating that the second imaging device is in an idle state in an abnormal state can be controlled to be transmitted to the manager terminal.

When it is confirmed in step S807 that the second imaging equipment is classified as an overload state, and it is confirmed that the overload state does not last longer than the reference time in step S809, the device provides a status notification message indicating that the second imaging equipment is in an overload state from an abnormal state can be controlled to be transmitted to the manager terminal.

If the second imaging device is classified as a replacement target device in step S810, the device displays a notification message indicating that the second imaging device needs to be replaced because the duration of the idle state of the second imaging device is long. It can be controlled to be transmitted to the terminal.

If the second imaging equipment is classified as the equipment to be repaired in step S811, the device sends a notification message to the manager terminal informing that the second photographing equipment needs to be repaired because the duration of the overload state of the second photographing equipment is long. transmission can be controlled.

9 is a diagram for explaining a first artificial neural network according to an embodiment.

The first artificial neural network 900 according to an embodiment receives a first input signal generated by encoding a first image that is an original image of a product photographed with a camera of a photographing equipment connected to the device as an input, and Information indicating the type can be output.

Encoding according to an embodiment may be performed by storing color information for each pixel of an image in the form of a digitized data sheet, and the color information may include RGB color, brightness information, and saturation information of one pixel. , but not limited thereto.

According to an embodiment, the first artificial neural network 900 is composed of a feature extraction neural network 910 and a classification neural network 920, and the feature extraction neural network 910 performs a task of separating a product and a background from an image. In addition, the classification neural network 920 may perform an operation of determining whether the product in the image belongs to which category.

The method in which the feature extraction neural network 910 distinguishes the product from the background is that the change in each value of color information from the data sheet of the first input signal encoding the image is 30% in 6 or more of 8 pixels including one pixel A bundle of pixels detected as having an abnormal change may be used as the boundary between the product and the background, but is not limited thereto.

The feature extraction neural network 910 stacks the input signal by sequentially stacking a convolutional layer and a pooling layer. The convolution layer includes a convolution operation, a convolution filter, and an activation function. The calculation of the convolution filter is adjusted according to the matrix size of the target input, but a 9X9 matrix is generally used. The activation function generally uses, but is not limited to, a ReLU function, a sigmoid function, and a tanh function. The pooling layer is a layer that reduces the size of the input matrix, and uses a method of extracting representative values by tying pixels in a specific area. In general, the average value or the maximum value is often used for the calculation of the pooling layer, but is not limited thereto. The operation is performed using a square matrix, usually a 9x9 matrix. The convolutional layer and the pooling layer are repeated alternately until the corresponding input becomes small enough while maintaining the difference.

The classification neural network 920 checks the shape of the product separated from the background through the feature extraction neural network 910, and checks whether the product in the image belongs to which category by checking whether it is similar to the representative product for each predefined category can figure out Information stored in the database of the device may be used for product comparison.

The classification neural network 920 prioritizes the task of identifying the category to which the product belongs, and may also grasp the shape and size of the product.

The classification neural network 920 has a hidden layer and an output layer, and is composed of five or less hidden layers, including a total of 50 or less hidden layer nodes, and the activation function of the hidden layer is a ReLU function and a sigmoid function. and tanh functions, but is not limited thereto.

The classification neural network 920 may include only one output layer node in total.

The output of the classification neural network 920 is an output value for the type of product, and may indicate to which category the product type belongs. For example, when the output value is 1, it indicates that the type of product is a category on clothes, when the output value is 2, it indicates that the type of product is a category under clothes, and when the output value is 3, it indicates that the type of product is It may indicate that it is a shoe category, and when the output value is 4, it may indicate that the type of product is a bag category.

According to an embodiment, the first artificial neural network 900 receives the first learning signal generated by the corrected correct answer input by the user when the user discovers a problem in the output according to the first artificial neural network 900 to learn. can The problem of output according to the first artificial neural network 900 may mean a case in which output values indicating different categories of products are output.

The first learning signal according to an embodiment is created based on the error between the correct answer and the output value, and in some cases, SGD using delta, a batch method, or a method following a backpropagation algorithm may be used. The first artificial neural network 900 performs learning by modifying the existing weights according to the first learning signal, and may use momentum in some cases. A cost function can be used to calculate the error, and a cross entropy function can be used as the cost function. Hereinafter, the learning contents of the first artificial neural network 900 will be described with reference to FIG. 10 .

10 is a diagram for explaining a method of learning a first artificial neural network according to an embodiment.

According to an embodiment, the learning apparatus may train the first artificial neural network 900 . The learning device may be a separate entity other than the device, but is not limited thereto.

According to an embodiment, the first artificial neural network 900 includes an input layer to which training samples are input and an output layer for outputting training outputs, and can be learned based on a difference between the training outputs and the first labels. have. Here, the first labels may be defined based on a representative image for each type of product. The first artificial neural network 900 is connected to a group of a plurality of nodes, and is defined by weights between the connected nodes and an activation function that activates the nodes.

The learning apparatus may train the first artificial neural network 900 by using a Gradient Decent (GD) technique or a Stochastic Gradient Descent (SGD) technique. The learning apparatus may use a loss function designed by the outputs and labels of the first artificial neural network 900 .

The learning apparatus may calculate a training error using a predefined loss function. The loss function may be predefined with a label, an output, and a parameter as input variables, where the parameter may be set by weights in the first artificial neural network 900 . For example, the loss function may be designed in a Mean Square Error (MSE) form, an entropy form, or the like, and various techniques or methods may be employed in an embodiment in which the loss function is designed.

The learning apparatus may find weights affecting the training error by using a backpropagation technique. Here, the weights are relationships between nodes in the first artificial neural network 900 . The learning apparatus may use the SGD technique using labels and outputs to optimize the weights found through the backpropagation technique. For example, the learning apparatus may update the weights of the loss function defined based on the labels, outputs, and weights using the SGD technique.

According to an embodiment, the learning apparatus may obtain the labeled training product images 1001 from the database. The learning apparatus may obtain pre-labeled information on each of the product images 1001 set for each type of product, and the product images 1001 may be labeled according to the pre-classified product information.

According to an embodiment, the learning device may obtain 1000 labeled training product images 1001 , and generate first training product vectors 1002 based on the labeled training product images 1001 . can Various schemes may be employed for extracting the first training product vectors 1002 .

According to an embodiment, the learning apparatus may obtain the first training outputs 1003 by applying the first training product vectors 1002 to the first artificial neural network 900 . The learning apparatus may train the first artificial neural network 900 based on the first training outputs 1003 and the first labels 1004 . The learning apparatus calculates training errors corresponding to the first training outputs 1003 , and optimizes the connection relationship of nodes in the first artificial neural network 900 to minimize the training errors to construct the first artificial neural network 900 . can learn The device may identify the type of product in the image using the first artificial neural network 900 on which the learning has been completed.

11 is an exemplary diagram of a configuration of an apparatus according to an embodiment.

According to one embodiment, the device 1100 includes a processor 1110 and a memory 1120 . The processor 1110 may include at least one of the devices described above with reference to FIGS. 1 to 10 , or perform at least one method described above with reference to FIGS. 1 to 10 . A person or group using the apparatus 1100 may provide a service related to some or all of the methods described above with reference to FIGS. 1 to 10 .

The memory 1120 may store information related to the above-described methods or a program for implementing the above-described methods. The memory 1120 may be a volatile memory or a non-volatile memory.

The processor 1110 may execute a program and control the device 1100 . The code of the program executed by the processor 1110 may be stored in the memory 1120 . The device 1100 may be connected to an external device (eg, a personal computer or a network) through an input/output device (not shown), and may exchange data through wired/wireless communication.

The device 1100 may be used to train an artificial neural network or use a learned artificial neural network. The memory 1120 may include a learning or learned artificial neural network. The processor 1110 may learn or execute an artificial neural network algorithm stored in the memory 1120 . The apparatus 1100 for learning the artificial neural network and the apparatus 1100 for using the learned artificial neural network may be the same or may be separate.

The embodiments described above may be implemented by a hardware component, a software component, and/or a combination of a hardware component and a software component. For example, the apparatus, methods and components described in the embodiments may include, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate (FPGA). array), a programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions. The processing device may execute an operating system (OS) and one or more software applications running on the operating system. The processing device may also access, store, manipulate, process, and generate data in response to execution of the software. For convenience of understanding, although one processing device is sometimes described as being used, one of ordinary skill in the art will recognize that the processing device includes a plurality of processing elements and/or a plurality of types of processing elements. It can be seen that can include For example, the processing device may include a plurality of processors or one processor and one controller. Other processing configurations are also possible, such as parallel processors.

The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the medium may be specially designed and configured for the embodiment, or may be known and available to those skilled in the art of computer software. Examples of the computer-readable recording medium include magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic such as floppy disks. - includes magneto-optical media, and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

The software may comprise a computer program, code, instructions, or a combination of one or more thereof, which configures a processing device to operate as desired or is independently or collectively processed You can command the device. The software and/or data may be any kind of machine, component, physical device, virtual equipment, computer storage medium or device, to be interpreted by or to provide instructions or data to the processing device. , or may be permanently or temporarily embody in a transmitted signal wave. The software may be distributed over networked computer systems, and stored or executed in a distributed manner. Software and data may be stored in one or more computer-readable recording media.

As described above, although the embodiments have been described with reference to the limited drawings, those skilled in the art may apply various technical modifications and variations based on the above. For example, the described techniques are performed in a different order than the described method, and/or the described components of the system, structure, apparatus, circuit, etc. are combined or combined in a different form than the described method, or other components Or substituted or substituted by equivalents may achieve an appropriate result.

Therefore, other implementations, other embodiments, and equivalents to the claims are also within the scope of the following claims.

Claims (3)

A method of processing an image for a three-dimensional cut of a product, performed by an apparatus, comprising:
when photographing is performed on a mannequin or a product worn by a person, acquiring a first image that is an original image of the product from photographing equipment;
generating a first input signal by encoding the first image;
inputting the first input signal to a first artificial neural network, and obtaining a first output signal based on a result of the input of the first artificial neural network;
identifying a first category that is a type of the product based on the first output signal;
comparing the representative image of the first category with the first image for each part, and setting a correction value for each part of the product;
generating a second image in which the fit of the product is corrected by performing a pixel fluidization operation for each part in the first image based on the correction value;
analyzing the second image to distinguish a product area and a background area occupied by the product in the second image;
deleting the background area from the second image to generate a third image that is an image composed of only the product area;
recognizing an outline of the product in the third image, and dividing the outline of the product into sections preset in the first category;
classifying each of the plurality of lines divided for each section in the outline of the product into a straight line and a curved line;
Recognizes each of a plurality of points located at a point where each of the plurality of lines is connected, and a first type in which a straight line and a straight line are connected to each of the plurality of points, a second type in which a straight line and a curved line are connected, and a curve and a straight line are connected to each of the plurality of points Classifying into a third type that becomes a curve and a fourth type in which the curve is connected; and
Including the step of registering information about the order, coordinates, and type of each of the plurality of points as characteristic information of the product,
How to process three-dimensional cut images of products. delete According to claim 1,
The step of classifying each of the plurality of lines into straight lines and curves,
recognizing a first point and a second point located at both ends of a first line that is any one of the plurality of lines;
recognizing a third point located in any part of the middle other than both ends of the first line;
generating a second line connecting the first point and the third point with a straight line and a third line connecting the second point and the third point with a straight line;
checking whether a fourth line combining the second line and the third line matches the first line; and
When it is confirmed that the fourth line and the first line match, the first line is classified as a straight line, and when it is confirmed that the fourth line and the first line do not match, the first line is classified as a curve comprising steps,
How to process three-dimensional cut images of products.

Images