WO2023185779A1

WO2023185779A1 - Method for identifying article information in refrigerator

Info

Publication number: WO2023185779A1
Application number: PCT/CN2023/084216
Authority: WO
Inventors: 毕研华; 孔令磊
Original assignee: 青岛海尔电冰箱有限公司; 海尔智家股份有限公司
Priority date: 2022-03-29
Filing date: 2023-03-28
Publication date: 2023-10-05
Also published as: CN114882348A

Abstract

Disclosed in the present invention is a method for identifying article information in a refrigerator. The method comprises: acquiring an aerial view image, identifying a contour frame which is formed by intersecting contour lines of the aerial view image, and segmenting the contour frame into a plurality of detection frames, wherein each detection frame corresponds to one article; selecting a contour line of the contour frame, and identifying the orientation of the contour line in the contour frame; if a relative orientation of a central point of the detection frame and the contour line is the same as the orientation of the contour line in the contour frame, the article being a mirror image article; and if the relative orientation of the central point of the detection frame and the contour line is opposite to the orientation of the contour line in the contour frame, the article being a real article. By means of the present invention, the interference of a mirror image article can be eliminated, and the accuracy of an identification algorithm is thus improved.

Description

Method for identifying item information in refrigerator

This application claims the priority of a Chinese patent application with an application date of March 29, 2022, an application number of 202210323179.9, and an invention titled "Method for identifying item information in refrigerator, refrigerator and computer storage medium", the entire content of which is incorporated by reference. in this application.

Technical field

The invention relates to the field of refrigeration devices, and in particular, to a method for identifying item information in a refrigerator.

Background technique

With the development of intelligent appliances, identifying internal items has become a necessary function of smart refrigerators. Usually, in order to realize the identification of items stored in the refrigerator, one or more cameras are installed inside the refrigerator to collect photos of the items in the refrigerator. The collected photos are identified and processed and then sent to the user terminal for viewing by the user. Placing the camera above objects can avoid the problem of objects blocking each other and improve the accuracy of object information recognition.

However, items placed inside the bottle holder will produce shadows on the bottle holder glass cabinet. Photos of items collected by the existing camera installed in the middle of the bottle holder often have an object mirroring problem, making it easy to identify one item as two items. Seriously affects the recognition effect. In addition, for some large-capacity and high-height drinks placed inside the bottle holder or items in the bottle holder stacked to a certain height, it will cause problems such as difficulty in distinguishing between real items in the bottle holder and their mirror images, and making it impossible to accurately identify them.

Contents of the invention

The object of the present invention is to provide a method for identifying item information in a refrigerator.

In order to achieve one of the above-mentioned objects of the invention, one embodiment of the present invention provides a method for identifying item information in a refrigerator, which includes the steps:

Obtain a bird's-eye view of the space inside the refrigerator bottle holder;

Identify a contour frame composed of intersecting contour lines in the bird's-eye view image;

Segment multiple detection frames in the bird's-eye image, each of the detection frames corresponding to an item;

Select a contour line and identify the position of the contour line in the contour frame;

Identify the center point of the detection frame;

If the relative orientation of the center point of the detection frame and the contour line is the same as the orientation of the contour line in the contour frame, then the corresponding item is a mirrored item;

If the relative orientation of the center point of the detection frame and the contour line is opposite to the orientation of the contour line in the contour frame, the corresponding item is a real item.

As a further improvement of an embodiment of the present invention, "identifying the composition of intersecting contour lines in the bird's-eye view image The "outline box" specifically includes:

Using an edge detection algorithm, the contour lines intersecting the frosted areas of the bird's-eye image are calibrated;

The contour lines intersect to form a contour frame.

As a further improvement of an embodiment of the present invention, "segmenting multiple detection frames in the bird's-eye view image, each of the detection frames corresponding to an item" specifically includes:

Using the Yolo V4 target detection algorithm, the detection frame corresponding to each item in the bird's-eye view image is calibrated;

The objects in the bird's-eye view image include real objects placed in the bottle holder and mirror objects corresponding to the real objects.

As a further improvement of one embodiment of the present invention, before "identifying the center point of the detection frame", the following steps are also included:

Obtain the detection frames on both sides of the contour line.

As a further improvement of one embodiment of the present invention, the method specifically includes:

Select a contour line, the contour line is located above the contour frame;

Identify the center point of the detection frame;

If the center point of the detection frame is located above the contour line, the corresponding item is a mirrored item;

If the center point of the detection frame is located below the contour line, the corresponding item is a real item.

Select a contour line, the contour line is located below the contour frame;

Identify the center point of the detection frame;

If the center point of the detection frame is located above the contour line, the corresponding item is a real item;

If the center point of the detection frame is located below the contour line, the corresponding item is a mirrored item.

In order to achieve one of the above objects, an embodiment of the present invention also provides another method for identifying item information in a refrigerator, including the steps:

Obtain a bird's-eye view of the space inside the refrigerator bottle holder;

If the ratio of the overlapping area of the detection frame and the outline frame to the area of the detection frame is less than the set threshold, the corresponding item is a mirrored item;

If the overlap area of the detection frame and the outline frame and the area of the detection frame are much greater than the set threshold, the corresponding item is a real item.

As a further improvement of an embodiment of the present invention, "identifying a contour frame composed of intersecting contour lines in the bird's-eye view image" specifically includes:

The contour lines intersect to form a contour frame.

As a further improvement of an embodiment of the present invention, the set threshold is specifically 5% to 10%.

Compared with the existing technology, the present invention provides a refrigerator in which the lower end of the wall of the bottle holder and the bottom of the bottle holder are frosted. By obtaining a bird's-eye view of the space inside the bottle holder of the refrigerator, the edge detection algorithm and the Yolo V4 target detection algorithm are used to calibrate the frosting. The contour frame of the area and the detection frame corresponding to the object in the image, obtain a contour line and identify the position of the contour line in the contour frame and identify the relative position of the center point of the object detection frame and the contour line, compare whether the two are the same to judge the bird's-eye view image Whether the items detected in the bottle are real items or mirrored items, this can eliminate the interference of mirrored items and improve the accuracy of the recognition algorithm; in addition, when some drinks with a higher height are placed in the bottle holder or the items in the bottle holder are stacked to a certain Height, because the lower end of the bottle holder wall is frosted, it can well avoid the problem that the center point of the object detection frame falls outside the outline frame, causing misjudgment by the recognition algorithm.

Description of drawings

Figure 1 is a schematic flowchart of a method for identifying item information in a refrigerator in an embodiment of the present invention.

Figure 2a is a schematic diagram of the space inside the bottle holder of the refrigerator in Embodiment 1 of the present invention (except for the outline for illustration and the items on adjacent sides, the rest of the content in the figure is omitted).

Figure 2b is a schematic flowchart of the method for identifying item information in the refrigerator in Embodiment 1 of the present invention.

Figure 3a is a schematic diagram of the space inside the bottle holder of the refrigerator in Embodiment 2 of the present invention (except for the outline for illustration and the items on adjacent sides, the rest of the content in the figure is omitted).

Figure 3b is a schematic flowchart of the method for identifying item information in the refrigerator in Embodiment 2 of the present invention.

Figure 4a is a schematic diagram of the space inside the bottle holder of the refrigerator in Embodiment 3 of the present invention (except for the outline for illustration and the items on adjacent sides, the rest of the content in the figure is omitted).

Figure 4b is a schematic flowchart of the method for identifying item information in the refrigerator in Embodiment 3 of the present invention.

Figure 5a is a schematic diagram of the space inside the bottle holder of the refrigerator in Embodiment 4 of the present invention (except for the outline for illustration and the items on adjacent sides, the rest of the content in the figure is omitted).

Figure 5b is a schematic flowchart of the method for identifying item information in the refrigerator in Embodiment 4 of the present invention.

FIG. 6 is a schematic diagram of the space inside the bottle holder of the refrigerator taken in another embodiment of the present invention.

Figure 7 is a schematic flowchart of a method for identifying item information in a refrigerator in another embodiment of the present invention.

Detailed ways

In order to make the purpose, technical solutions and advantages of the present application clearer, the technical solutions of the present application will be clearly and completely described below in conjunction with the specific implementation modes of the present application and the corresponding drawings. Obviously, the described embodiments are only some of the embodiments of the present application, but not all of them. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of this application.

Embodiments of the present invention are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements with the same or similar functions. The embodiments described below with reference to the drawings are exemplary and are only used to explain the present invention and are not to be construed as limitations of the present invention.

For convenience of explanation, this article uses terms indicating relative positions in space, such as "upper", "lower", "back", "front", etc., to describe one unit or feature shown in the drawings relative to another. A relationship between a unit or a feature. Spatially relative terms may refer to different orientations of the device in use or operation in addition to the orientation illustrated in the figures. For example, if the device in the diagram is turned over, elements described as "below" or "above" other elements or features would then be oriented "below" or "above" the other elements or features. Thus, the exemplary term "below" may encompass both spatial orientations, below and above.

As shown in Figure 1, the present invention provides a method for identifying item information in a refrigerator, which includes the steps:

S1: Obtain a bird's-eye view of the space inside the refrigerator bottle holder.

In this embodiment, a camera is installed on the top of the bottle holder, and the objects are photographed from above to obtain a bird's-eye view image. In order to avoid mutual occlusion between objects in the bottle holder, which is not conducive to the detection and identification of object information, the present invention preferably uses the camera Located at the middle position of the top of the bottle holder.

Of course, in other embodiments of the present invention, two or more cameras may also be provided in the bottle holder to obtain two or more bird's-eye images of the space inside the bottle holder of the refrigerator taken from different angles.

S2: Identify a contour frame composed of intersecting contour lines in the bird's-eye view image.

In the specific embodiment of the present invention, at least one wall of the bottle holder in the refrigerator is a glass surface, which is prone to mirror interference. Specifically, the area of the glass surface close to the bottom of the bottle holder is frosted to ensure large-capacity and high-height items. While the center point of the detection frame falls within the outline frame, it does not prevent the user from observing the inside of the bottle holder from outside the glass surface of the bottle holder. Preferably, in the present invention, the 10cm wide area at the lower end of the glass surface is frosted, and the bottom of the bottle holder is also frosted. Matte finish. Since there is an obvious difference in light transmittance between frosted areas and non-frosted areas, areas such as the bottom of the bottle holder and the wall of the bottle holder captured by the camera usually have obvious pixel grayscale differences in the image, which can be detected and identified. Edge outline.

Furthermore, the present invention uses an edge detection algorithm to identify the edge contours of the bottle holder, and calibrate the contour lines intersecting the frosted areas of the bird's-eye image. The intersection of the contour lines forms an outline frame. For easier explanation, this embodiment uses a refrigerator bottle Taking the bottom end of the four wall surfaces of the seat as an example with a 10cm wide frosted area, the calibrated outline frame 1 is a rectangular structure as shown in Figures 2a, 3a, 4a and 5a. Here, edge detection algorithms can be used such as Roberts, Sobel, Currently common algorithms such as Prewitt, Canny, and Log are based on existing technologies and will not be described in detail here.

S3: Segment multiple detection frames in the bird's-eye view image, and each detection frame corresponds to an item.

In the non-frosted area of the bottle holder wall, the objects in the bottle holder often cast corresponding shadows in the non-frosted area, that is, mirror objects. Therefore, the objects in the bird's-eye view image captured in the present invention include objects placed on the bottle holder. Real objects in the seat and mirrored objects corresponding to said real objects.

Specifically, the items present in the bird's-eye view image are identified, and the Yolo V4 target detection algorithm is used to calibrate the detection frame 2 corresponding to each item in the bird's-eye view image. The specific content of the Yolo V4 target detection algorithm here is the existing technology, so I won’t go into details here.

S4: Select a contour line and identify the position of the contour line in the contour frame.

Specifically, the outline frame 1 is a square structure, which includes an outline line 11 located above the outline frame, an outline line 12 located below the outline frame, an outline line 13 located on the left side of the outline frame, and an outline line 13 located on the right side of the outline frame. Orientation contour line 14.

For example, in the embodiment of the present invention, the selected contour line may be contour line 11, contour line 12, contour line 13 or contour line 14.

S5: Identify the center point of the detection frame.

Specifically, before "identifying the center point of the detection frame", it also includes the step of: obtaining a detection frame adjacent to the contour line. For example, if the contour line 11 is selected in step S4, then the system only obtains the object detection frames on both sides adjacent to the contour line 11, and does not use this as a criterion for the object detection frames on other surfaces.

Whether the object identified in the bird's-eye view image is a real object or a mirror image is determined based on whether the relative orientation of the identified center point 3 of the detection frame and the contour line is the same as the orientation of the contour line in the contour frame 1 .

S61: If the relative orientation of the center point of the detection frame and the contour line is the same as the orientation of the contour line in the contour frame, then the corresponding item is a mirror image item.

S62: If the relative orientation of the center point of the detection frame and the contour line is opposite to the orientation of the contour line in the contour frame, then the corresponding item is a real item.

For example, if the contour line 11 is selected in step S4, then in the step "obtain detection frames on both sides adjacent to the contour line", only item a, item b, item c and as shown in Figure 2a are obtained. The detection frame of item d determines whether it is a real item or a mirrored item. For item a, the relative position between the center point 3a of the detection frame 2a and the selected contour line 11 is that the center point 3a of the detection frame 2a is located above the contour line 11, and the contour line 11 is located above the contour frame 1, Therefore, the relative orientation of the center point 3a of the detection frame 2a of item a and the contour line 11 is the same as the orientation of the contour line 11 in the contour frame 1, that is, step S61 is executed, and the system determines that item a is a mirrored item.

Similarly, for item b, item c and item d in Figure 2a, according to the above determination method, it can be effectively determined that item b is a real item, item c is a mirror item and item d is a real item.

What needs to be explained here is that for items with large capacity and high height, such as item d in Figure 2a, or bottles There are multiple items stacked together to a certain height in the holder (not shown in the figure), and the center point 3d of the detection frame is often higher than the edge line 4 of the bottom of the bottle holder. Therefore, in this application, the lower end of the glass surface of the bottle holder is The 10cm wide area is frosted, and the frosted area in the entire bird's-eye view image is identified, that is, the outline frame 1 of the entire frosted area is calibrated. This ensures that the center point of the detection frame 2 corresponding to such items with higher heights falls within the outline frame 1. At the same time, it will not affect the user's ability to observe the stacking of items in the bottle holder from the outside of the bottle holder.

Due to different selected contours, the present invention provides four specific identification methods for items in the bottle holder of the refrigerator, which are described in detail below with four embodiments.

Embodiment 1

In Embodiment 1, the contour line 11 is selected, as shown in Figure 2a, and the identification method (Figure 2b) includes the steps:

S4A1: Select a contour line 11, which is located on the upper edge of the contour frame 1.

S5A1: Identify the center point 3 of the detection frame.

S61A1: If the center point 3 of the detection frame is located above the contour line 11, the corresponding item is a mirrored item.

S62A1: If the center point 3 of the detection frame is located below the contour line 11, the corresponding item is a real item.

Similarly, before identifying the center point 3 of the detection frame, a step is also included: obtaining the detection frames 2 on both sides adjacent to the contour line 11 . As shown in Figure 2a, for item a in the figure, because the center point 3a of the detection frame is located above the contour line 11, it is judged that item a is a mirrored item; for item b in the figure, because the center point 3b of the detection frame is located above the contour line 11 Below, it is judged that item b is a real item; for item c in the picture, because the center point 3c of its detection frame is located above the contour line 11, it is judged that item c is a mirror item; for item d in the picture, because the center point of its detection frame 3d is located below the contour line 11, that is, item d is judged to be a real item.

In this embodiment, the object detection frames on adjacent sides of the contour line 11 obtained in the bird's-eye view can use the object information recognition method in Figure 2b to determine whether the object is a real object or a mirror object.

Embodiment 2

In the second embodiment, the contour line 12 is selected, as shown in Figure 3a. The identification method (Figure 3b) includes the steps:

S4B1: Select a contour line 12 located at the lower edge of the contour frame 1.

S5B1: Identify the center point 3 of the detection frame.

S61B1: If the center point 3 of the detection frame is located above the contour line 12, the corresponding item is a real item.

S62B1: If the center point 3 of the detection frame is located below the contour line 11, the corresponding item is a mirrored item.

Similarly, before identifying the center point 3 of the detection frame, a step is also included: obtaining the detection frames 2 on both sides adjacent to the contour line 12 . As shown in Figure 3a, for item e in the figure, since the center point 3e of its detection frame is located on the contour line 12 Above, it is judged that item e is a real item; for item f in the picture, since the center point 3f of the detection frame is located below the contour line 12, it is judged that item f is a mirror item.

In this embodiment, the object detection frames on adjacent sides of the contour line 12 obtained in the bird's-eye view can use the object information recognition method in Figure 3b to determine whether the object is a real object or a mirror object.

Embodiment 3

In the third embodiment, the contour line 13 is selected, as shown in Figure 4a, and the identification method (Figure 4b) includes the steps:

S4C1: Select a contour line 13, which is located on the left side of the contour frame 1.

S5C1: Identify the center point 3 of the detection frame.

S61C1: If the center point 3 of the detection frame is located to the left of the contour line 13, the corresponding item is a mirror image item.

S62C1: If the center point 3 of the detection frame is located to the right of the contour line 13, the corresponding item is a real item.

Similarly, before identifying the center point 3 of the detection frame, a step is also included: obtaining detection frames on both sides adjacent to the contour line 13 . As shown in Figure 4a, for item g in the figure, because the center point 3g of its detection frame is located to the left of the contour line 13, it is judged that item g is a mirrored item; for item h in the figure, because the center point 3h of its detection frame is located on the outline To the right of line 13, it is judged that item h is a real item.

In this embodiment, the object detection frames on adjacent sides of the contour line 13 obtained in the bird's-eye view can use the object information recognition method in Figure 4b to determine whether the object is a real object or a mirror image.

Embodiment 4

In the fourth embodiment, the contour line 14 is selected, as shown in Figure 5a. The identification method (Figure 5b) includes the steps:

S4D1: Select a contour line 14, which is located on the right side of the contour frame 1.

S5D1: Identify the center point 3 of the detection frame. .

S61D1: If the center point 3 of the detection frame is located to the left of the contour line 14, the corresponding item is a real item.

S62D1: If the center point 3 of the detection frame is located to the right of the contour line 14, the corresponding item is a mirrored item.

Similarly, before identifying the center point of the detection frame, a step is also included: obtaining detection frames on both sides adjacent to the contour line 14 . As shown in Figure 5a, for item i in the figure, since the center point 3i of the detection frame is located on the left side of the contour line 14, it is judged that item i is a real item; for item j in the figure, because the center point 3j of the detection frame is located on the To the right of the contour line 14, it is determined that item j is a mirror image item.

In this embodiment, the object detection frames on adjacent sides of the contour line 14 obtained in the bird's-eye view can use the object information recognition method in Figure 5b to determine whether the object is a real object or a mirror image.

Furthermore, considering that recognition errors may occur during the actual use of the recognition algorithm, it is better to For example, in Figure 2a, items a and item c are too close to the contour line 11, causing part of the detection frame to calibrate item a and item c to fall into the outline frame 1. As shown in Figure 6, the detection frames 2a and 3a are not consistent with the outline frame. 1 has an overlapping area, resulting in the center points 3a and 3b of the contour frame being less than 25 pixels away from the contour line 11. Therefore, the present invention proposes another method for identifying item information in the refrigerator based on possible algorithm errors, as shown in Figure 7, including the steps:

S41': If the ratio of the overlapping area of the detection frame and the outline frame to the area of the detection frame is less than the set threshold, the corresponding item is a mirrored item.

S42': If the overlap area of the detection frame and the outline frame and the area of the detection frame are much greater than the set threshold, the corresponding item is a real item.

In this embodiment, the set threshold is 5% to 10%.

The invention also provides a refrigerator, which includes a bottle holder, a camera, a memory and a processor.

Further, the bottle holder includes a top, a bottom and a wall of the bottle holder, and the camera is located at the middle position of the top of the bottle holder, vertically downward.

The bottom of the bottle holder is used to place items. At least one side of the wall of the bottle holder is a glass surface. A 10cm wide area of the glass surface close to the bottom of the bottle holder is frosted. The bottom of the bottle holder is also frosted.

The memory stores a computer program that can be run on the processor. When the processor executes the program, the steps of the above method for identifying item information in the refrigerator are implemented.

The present invention also provides a computer storage medium in which a computer program is stored, and when the computer program is run, the above steps of the method for identifying item information in the refrigerator are executed on the device where the computer storage medium is located.

In summary, the present invention provides a refrigerator in which the lower end of the bottle holder wall and the bottom of the bottle holder are frosted. By obtaining a bird's-eye view image of the space inside the refrigerator bottle holder, the edge detection algorithm and the Yolo V4 target detection algorithm are used to calibrate the frosted area. The contour frame and the detection frame corresponding to the object in the image, obtain a contour line and identify the position of the contour line in the contour frame and identify the relative position of the center point of the object detection frame and the contour line, compare whether the two are the same to determine the detection in the bird's-eye view image Whether the received item is a real item or a mirrored item, this eliminates the interference of mirrored items and improves the accuracy of the recognition algorithm; in addition, when some drinks with a higher height are placed in the bottle holder or the items in the bottle holder are stacked to a certain height, Since the lower end of the bottle holder wall is frosted, it can well avoid the problem that the center point of the object detection frame falls outside the outline frame, causing misjudgment by the recognition algorithm.

It should be understood that although this specification is described in terms of implementations, not each implementation only contains an independent technical solution. This description of the specification is only for the sake of clarity. Persons skilled in the art should take the specification as a whole and understand each individual solution. The technical solutions in the embodiments can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

The series of detailed descriptions listed above are only specific descriptions of feasible implementations of the present invention and are not intended to limit the scope of protection of the present invention. Any equivalent implementations or changes that do not depart from the technical spirit of the present invention are All should be included in the protection scope of the present invention.

Claims

A method for identifying item information in a refrigerator, which is characterized by including the steps:

Obtain a bird's-eye view of the space inside the refrigerator bottle holder;

Identify a contour frame composed of intersecting contour lines in the bird's-eye view image;

Segment multiple detection frames in the bird's-eye image, each of the detection frames corresponding to an item;

Select a contour line and identify the position of the contour line in the contour frame;

Identify the center point of the detection frame;

If the relative orientation of the center point of the detection frame and the contour line is the same as the orientation of the contour line in the contour frame, then the corresponding item is a mirrored item;

If the relative orientation of the center point of the detection frame and the contour line is opposite to the orientation of the contour line in the contour frame, the corresponding item is a real item.
The method for identifying item information in a refrigerator according to claim 1, wherein "identifying a contour frame composed of intersecting contour lines in the bird's-eye view image" specifically includes:

Using an edge detection algorithm, the contour lines intersecting the frosted areas of the bird's-eye image are calibrated;

The contour lines intersect to form a contour frame.
The method for identifying item information in a refrigerator according to claim 2, characterized in that "segmenting a plurality of detection frames in the bird's-eye view image, each of the detection frames corresponding to an item" specifically includes:

Using the Yolo V4 target detection algorithm, the detection frame corresponding to each item in the bird's-eye view image is calibrated;

The objects in the bird's-eye view image include real objects placed in the bottle holder and mirror objects corresponding to the real objects.
The method for identifying item information in a refrigerator according to claim 3, characterized in that before "identifying the center point of the detection frame", it further includes the step of:

Obtain the detection frames on both sides of the contour line.
The method for identifying item information in a refrigerator according to claim 4, wherein the method specifically includes:

Select a contour line, the contour line is located above the contour frame;

Identify the center point of the detection frame;

If the center point of the detection frame is located above the contour line, the corresponding item is a mirrored item;

If the center point of the detection frame is located below the contour line, the corresponding item is a real item.
The method for identifying item information in a refrigerator according to claim 5, wherein the method specifically includes:

Select a contour line, the contour line is located below the contour frame;

Identify the center point of the detection frame;

If the center point of the detection frame is located above the contour line, the corresponding item is a real item;

If the center point of the detection frame is located below the contour line, the corresponding item is a mirrored item.
A method for identifying item information in a refrigerator, which is characterized by including the steps:

Obtain a bird's-eye view of the space inside the refrigerator bottle holder;

Identify a contour frame composed of intersecting contour lines in the bird's-eye view image;

Segment multiple detection frames in the bird's-eye image, each of the detection frames corresponding to an item;

If the ratio of the overlapping area of the detection frame and the outline frame to the area of the detection frame is less than the set threshold, the corresponding item is a mirrored item;

If the overlap area of the detection frame and the outline frame and the area of the detection frame are much greater than the set threshold, the corresponding item is a real item.
The method for identifying item information in a refrigerator according to claim 7, wherein "identifying a contour frame composed of intersecting contour lines in the bird's-eye view image" specifically includes:

Using an edge detection algorithm, the contour lines intersecting the frosted areas of the bird's-eye image are calibrated;

The contour lines intersect to form a contour frame.
The method for identifying item information in a refrigerator according to claim 8, characterized in that "segmenting a plurality of detection frames in the bird's-eye view image, each of the detection frames corresponding to an item" specifically includes:

Using the Yolo V4 target detection algorithm, the detection frame corresponding to each item in the bird's-eye view image is calibrated;

The objects in the bird's-eye view image include real objects placed in the bottle holder and mirror objects corresponding to the real objects.
The method for identifying item information in a refrigerator according to claim 7, wherein the set threshold is specifically 5% to 10%.