Method for identifying information of items in refrigerator, refrigerator and computer storage medium

Abstract

The present invention discloses a method for identifying information of items in a refrigerator, a refrigerator and a computer storage medium, the method comprising the steps of: obtaining a bird's-eye view image of the space inside the refrigerator bottle holder; increasing the brightness of the bird's-eye view image to 1.5 to 2.5 times of the original brightness; selecting an outer area of ​​the bird's-eye view image, and determining the pixel similarity range according to the pixel value data in the area; performing pixel similarity calibration on all pixels in the bird's-eye view image according to the pixel similarity range to form a calibration set, the calibration set including all pixel data in the bird's-eye view image that fall within the pixel similarity range; excluding all pixels in the calibration set to obtain image information of real items in the bottle holder. The present invention utilizes pixel calibration to remove the mirror image area of ​​the refrigerator bottle holder bird's-eye view, thereby eliminating the interference of mirror image items, obtaining image information of real items in the bottle holder, and improving the accuracy of the recognition algorithm.

Description

Technical Field

[0001] The present invention relates to the field of refrigeration devices, and in particular to a method for identifying information of items in a refrigerator, a refrigerator and a computer storage medium. Background Art

[0002] With the development of intelligent electrical appliances, identifying internal items has become a necessary function for 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, and the collected photos are sent to the user terminal after identification processing for users to view. Placing the camera above the items can avoid the problem of items blocking each other and improve the accuracy of item information recognition.

[0003] However, objects placed inside the bottle holder will cast shadows on the bottle holder glass cabinet. The existing camera is installed in the middle of the bottle holder, and the photos of objects collected often have object mirroring problems, which can easily identify one object as two objects, seriously affecting the recognition effect.

[0004] The existing method is to manually cut off the mirror area in the refrigerator bottle holder by fixing the ROI (Region of Interest), but this method will cause the ROI area to be too large or too small, which is not accurate enough and wastes manpower; or the method of using frosted material on the wall of the bottle holder where the objects are placed to eliminate the influence of the object's mirror image is adopted, but this method not only increases the production process, but also affects the appearance of the whole machine. Summary of the invention

[0005] The object of the present invention is to provide a method for identifying information of items in a refrigerator, a refrigerator and a computer storage medium.

[0006] To achieve one of the above-mentioned purposes of the invention, an embodiment of the present invention provides a method for identifying information of items in a refrigerator, comprising the steps of:

[0007] Get a bird's-eye view of the space inside the refrigerator bottle holder;

[0008] Increase the brightness of the bird's-eye view image to 1.5 to 2.5 times of the original brightness;

[0009] Selecting an outer area of ​​the bird's-eye view image, and determining the pixel similarity range according to pixel value data in the area;

[0010] Performing pixel similarity calibration on all pixels in the bird's-eye view image according to the pixel similarity range to form a calibration set, wherein the calibration set includes all pixel data in the bird's-eye view image that fall within the pixel similarity range;

[0011] All pixel points in the calibration set are excluded to obtain image information of the real object in the bottle holder.

[0012] As a further improvement of the present invention, “obtaining a bird's-eye view image of the space inside the refrigerator bottle holder” specifically includes:

[0013] A camera arranged in the middle of the top of the bottle seat is controlled to capture a bird's-eye view image of the space inside the bottle seat.

[0014] As a further improvement of the present invention, “selecting the outer area of ​​the bird's-eye view image” specifically includes:

[0015] Select any three outermost areas of the bird's-eye view image, where the three areas are spaced a certain distance apart;

[0016] The three areas include a first area, a second area and a third area, wherein:

[0017] The first area and the third area are located at two outermost ends of the bird's-eye view image, respectively, and the second area is located at the outermost middle position of the bird's-eye view image.

[0018] As a further improvement of the present invention, “determining the pixel similarity range according to the pixel value data in the region” specifically includes the steps of:

[0019] Obtain pixel values ​​within any area of ​​the three areas, and calculate an average value of the pixel values;

[0020] Obtain pixel values ​​within the other two regions, and calculate the maximum and minimum values ​​of the pixels within the two regions;

[0021] The pixel similarity range is determined according to the average value of pixels in the one region and the maximum and minimum values ​​of pixels in the other two regions.

[0022] As a further improvement of the present invention, “performing pixel similarity calibration on all pixels in the bird's-eye view image according to the pixel similarity range to form a calibration set” specifically includes the steps of:

[0023] Obtaining pixel value data of all pixels in the bird's-eye view image;

[0024] According to the pixel similarity range, a FloodFill algorithm is used to perform pixel similarity calibration on all pixels in the bird's-eye view image;

[0025] Form a calibration set.

[0026] As a further improvement of the present invention, “excluding all pixels in the calibration set” specifically includes:

[0027] Obtain any pixel value within the pixel similarity range, covering the calibration set.

[0028] As a further improvement of the present invention, the pixel value data specifically includes R, G, and B three-channel pixel value data.

[0029] As a further improvement of the present invention, the method further comprises the steps of:

[0030] The number of real objects in the bird's-eye view image is counted.

[0031] An embodiment of the present invention provides a refrigerator, comprising: a bottle holder, a camera, a memory and a processor, wherein:

[0032] The bottle seat comprises a top, a bottom and a wall of the bottle seat, and the camera is arranged in the middle of the top of the bottle seat, and is arranged vertically downward;

[0033] The bottom of the bottle seat is used for placing items;

[0034] At least one side of the bottle seat wall is a glass surface;

[0035] The memory stores a computer program that can be run on the processor, and when the processor executes the program, the steps of the method for identifying information of items in a refrigerator as described in any one of the above items are implemented.

[0036] An embodiment of the present invention further provides a computer storage medium, which stores a computer program, wherein when the computer program is executed, the device where the computer storage medium is located executes the steps of the method for identifying information of items in a refrigerator as described in any one of the above items.

[0037] Compared with the prior art, the present invention improves the brightness of the bird's-eye view of the space inside the bottle holder, calculates the pixel similarity range based on the pixel value data of the outer area, and then uses the FloodFill algorithm to calibrate the pixel similarity of all pixels in the image to form a calibration set. The pixel point data in the calibration set can be excluded, that is, the mirror area of ​​the bird's-eye view of the refrigerator bottle holder is removed, thereby eliminating the interference of the mirrored objects, obtaining the image information of the real objects in the bottle holder, and improving the accuracy of the recognition algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a flow chart of a method for identifying information of items in a refrigerator in one embodiment of the present invention.

[0039] Figure 2 is a flowchart of the steps of a method for determining a pixel similarity range.

[0040] FIG. 3( a ) and FIG. 3( b ) are schematic diagrams of photographing the space inside the bottle holder of the refrigerator in the first embodiment of the present invention.

[0041] FIG. 4( a ) and FIG. 4( b ) are schematic diagrams of photographing the space inside the bottle holder of the refrigerator in the second embodiment of the present invention.

[0042] FIG. 5( a ) and FIG. 5( b ) are schematic diagrams of photographing the space inside the bottle holder of the refrigerator in the third embodiment of the present invention.

[0043] 6(a) and 6(b) are schematic diagrams of photographing the space inside the bottle holder of the refrigerator in the fourth embodiment of the present invention. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solutions and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in combination with the specific implementation methods of this application and the corresponding drawings. Obviously, the described implementation methods are only part of the implementation methods of this application, not all of the implementation methods. Based on the implementation methods in this application, all other implementation methods obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0045] The embodiments of the present invention are described in detail below, and 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 having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0046] For ease of explanation, this document uses terms that represent relative positions in space for description, such as "above", "below", "back", "front", etc., to describe the relationship of one unit or feature shown in the drawings relative to another unit or feature. Terms of relative positions in space can include different orientations of the device in use or operation other than the orientation shown in the figure. For example, if the device in the figure is turned over, the unit described as being "below" or "above" other units or features will be located "below" or "above" the other units or features. Therefore, the exemplary term "below" can encompass both the spatial orientations of below and above.

[0047] like Figure 1 As shown, the present invention provides a method for identifying information of items in a refrigerator, comprising the steps of:

[0048] S1: Obtain a bird's-eye view image of the space inside the refrigerator bottle holder.

[0049] In this embodiment, the camera is arranged on the top of the bottle holder to photograph the objects from above to obtain a bird's-eye view image. In order to avoid mutual occlusion between the objects in the bottle holder, which is not conducive to the detection and identification of the object information, the present invention preferably arranges the camera in the middle position of the top of the bottle holder.

[0050] Of course, in some other embodiments of the present invention, two or more cameras may be arranged in the bottle holder to obtain two or more bird's-eye views of the space inside the refrigerator bottle holder taken from different angles.

[0051] S2: increasing the brightness of the bird's-eye view image to 1.5 to 2.5 times of the original brightness.

[0052] When the wall of the bottle holder is not made of frosted material, the objects in the bottle holder will often project corresponding shadows on the wall of the bottle holder, that is, their mirror images. Therefore, the objects in the bird's-eye view image obtained in the present invention include real objects placed in the bottle holder and mirror images corresponding to the real objects.

[0053] Since the bottom of the bottle seat, the object in the bottle seat and the wall of the bottle seat are made of different materials, the bottom of the bottle seat, the object and the wall of the bottle seat captured by the camera usually have obvious pixel grayscale differences in the image. At this time, increasing the brightness of the bird's-eye view image will fade the image of the object on the wall of the bottle seat.

[0054] Specifically, the brightness of the bird's-eye view image is adjusted to 1.5 to 2.5 times of the original brightness. If the brightness is adjusted too low, the mirror image of the object in the bottle holder on the wall of the bottle holder cannot be weakened. If the brightness is adjusted too high, the real object in the bottle holder will be exposed, affecting the accuracy of the subsequent elimination of the mirror image area. In one embodiment of the present invention, the brightness of the bird's-eye view image is preferably adjusted to twice the brightness of the original image.

[0055] S3: Selecting an outer area of ​​the bird's-eye view image, and determining the pixel similarity range according to pixel value data in the area.

[0056] In one embodiment of the present invention, Figure 2 As shown, step S3 specifically includes the following steps:

[0057] S31: Select any three outermost areas of the bird's-eye view image, where the three areas are spaced apart by a certain distance.

[0058] Specifically, the three regions include a first region, a second region, and a third region. In one embodiment of the present invention, the first region and the third region are located at the outermost ends of the bird's-eye view image, respectively, and the second region is located in the outermost middle region of the bird's-eye view image. Of course, in other embodiments, the specific positions of the three regions may not be limited, and the three regions are only required to be spaced a certain distance apart to reduce the error in the subsequent calculation of the pixel similarity range due to the selected regions being too close.

[0059] Furthermore, in a specific embodiment of the present invention, the selected area range is preferably a 20×20 pixel range. Of course, a larger pixel range or a smaller pixel range can also be used for subsequent calculation operations. Therefore, changes in the specific range of the selected bird's-eye view image calculation area are within the protection scope of the present invention.

[0060] S32: Obtain pixel values ​​within the range of any one of the three regions, and calculate an average value of the pixel values.

[0061] Specifically, the pixel values ​​obtained specifically include multiple groups of R, G, and B three-channel pixel value data, that is, the average values ​​of the R, G, and B three-channel pixel points obtained within any area are calculated, that is,

[0062] S33: Obtain pixel values ​​within the other two regions, and calculate the maximum and minimum values ​​of the pixels within the two regions.

[0063] Similarly, the pixel values ​​within the other two regions are obtained, including multiple sets of R, G, and B three-channel pixel value data, and the maximum and minimum values ​​of the multiple sets of data are calculated, that is, ((R max , R min ), (G max , G min ), (B max , B min )).

[0064] In one embodiment of the present invention, the pixel points within the first area can be selected to calculate the pixel average value of the RGB three channels, and the pixel points within the second area and the third area can be selected to calculate the pixel maximum and minimum values ​​of the RGB three channels. Of course, in other embodiments of the present invention, the pixel points within the second area or the third area can be selected to calculate the pixel average value of the RGB three channels, and the pixel points within the other two channels can be selected to calculate the pixel maximum and minimum values ​​of the RGB three channels.

[0065] S34: Determine a pixel similarity range according to the average pixel value in the one region and the maximum and minimum pixel values ​​in the other two regions.

[0066] Specifically, according to the average value of the three channels R, G, and B of the pixel points within a region obtained in step S32 In step S33, the maximum and minimum values ​​of the three channels R, G, and B in the pixel data within the other two regions are obtained (R max , R min ), (G max , G min ), (B max , B min )), and get a set of RGB three-channel pixel similarity ranges, namely:

[0067]

[0068]

[0069]

[0070] Furthermore, the present invention is not limited to selecting only three areas outside the bird's-eye view for subsequent calculation to obtain the pixel similarity range, but four or more areas can also be selected, and the pixel average value is calculated from the pixel value data within any one of the areas, and the pixel maximum and minimum values ​​are calculated from the pixel value data within other areas to determine the final pixel similarity range.

[0071] S4: performing pixel similarity calibration on all pixels in the bird's-eye view image according to the pixel similarity range to form a calibration set, wherein the calibration set includes all pixel data in the bird's-eye view image that fall within the pixel similarity range.

[0072] Specifically, the pixel value data of all pixels in the bird's-eye view image are obtained, and it is determined whether the RGB three-channel values ​​of all pixels fall within the above-mentioned RGB three-channel pixel similarity range. If so, pixel calibration is performed. In one embodiment of the present invention, according to the pixel similarity range, the FloodFill algorithm is used to perform pixel similarity calibration on all pixels in the bird's-eye view image to form a calibration set. Here, the FloodFill algorithm is a prior art and will not be described in detail.

[0073] S5: Eliminate all pixel points in the calibration set to obtain image information of the real object in the bottle holder.

[0074] Specifically, any pixel value within the pixel similarity range, that is, any group of RGB values ​​within the above-mentioned RGB three-channel pixel similarity range is selected, and the calibration set is used as a mask. Any pixel value within the pixel similarity range is used to cover the pixel points in the calibration set to identify the accurate mirror area of ​​the bottle seat wall to the greatest extent.

[0075] Furthermore, since the pixel values ​​of the bottom of the bottle seat and the objects placed on the bottom of the bottle seat are greatly different from the pixel values ​​of the glass surface of the bottle seat, their pixel points will not fall into the calibration set. In this way, the mirror effect of the objects in the bottle seat on the wall of the bottle seat can be eliminated, and the image information of the real objects in the bottle seat can be obtained.

[0076] Since the refrigerator bottle seat includes a bottle seat bottom and four bottle seat walls, the present invention provides four specific methods for identifying objects in the bottle seat when the bottle seat walls at different positions of the bottle seat generate mirror images. The following four embodiments are specifically described.

[0077] Embodiment 1

[0078] In Embodiment 1, one upper side 1 in the bird's-eye view image of the bottle holder of the refrigerator is a glass surface, as shown in FIG3(a). The real object in the bottle holder has its mirror image formed on the upper side 1, which interferes with the final system's acquisition of information about the real object in the bottle holder. In this regard, using the above-mentioned object recognition method, in step S31, any three areas on the upper side of the bird's-eye view image are selected, preferably the upper left, upper middle, and upper right three areas, which are recorded as the first area 11, the second area 12, and the third area 13.

[0079] Preferably, the first area 11, the second area 12, and the third area 13 are all in the range of 20×20 pixels. The pixel data within the range of the first area 11 is obtained, and the pixel average value of the RGB three channels of the pixel points within the range is calculated, which is recorded as Get the maximum and minimum values ​​of the RGB channels of the pixels in the second area 12 and the third area 13, denoted as ((R 1max , R 1min ), (G 1max , G 1min ), (B 1max , B 1min )), thus obtaining a set of R1G1B1 three-channel pixel similarity ranges, namely:

[0080]

[0081]

[0082]

[0083] The RGB three-channel data of all pixels in the bird's-eye view image are obtained. According to the above pixel similarity range, the FloodFill algorithm is used to perform pixel similarity calibration on all pixels in the bird's-eye view image to form a calibration set. Then, any set of R1G1B1 values ​​is selected within the above pixel similarity range, and the calibration set is used as a mask to cover the pixels in the calibration set with any selected set of R1G1B1 values. This can identify the upper side 1 in the bird's-eye view image to the greatest extent, as shown in FIG3(b). All pixels in the calibration set can be excluded to obtain the image information of the real object in the bottle holder.

[0084] Furthermore, when there is a large object a in the bottle holder, such as a large volume of milk or beverage, in the bird's-eye view image, a portion of the large object a will fall on the upper side surface 1 due to its large volume, but since the pixel value of the object placed at the bottom of the bottle holder is greatly different from the pixel value of the glass surface of the bottle holder, the pixel points on the surface of the real object will not fall into the calibration set, nor will they be covered by the optional RGB pixel value, so when covering the pixel data in the upper side surface 1, it will not affect the large volume of the real object, and will not affect the image information of the real object in the bottle holder that is finally obtained. Similarly, the following three embodiments are all the same.

[0085] Embodiment 2

[0086] In the second embodiment, a lower side surface 2 in the bird's-eye view image of the bottle holder of the refrigerator is a glass surface, as shown in FIG4(a). The real object in the bottle holder has its mirror image formed on the lower side surface 2, which interferes with the final system's acquisition of the real object information in the bottle holder. In this regard, using the above-mentioned object recognition method, in step S31, any three areas at the bottom of the bird's-eye view image are selected, preferably the lower left, lower middle, and lower right three areas, which are recorded as the first area 21, the second area 22, and the third area 23.

[0087] Similarly, the first area 21, the second area 22, and the third area 23 are preferably in the range of 20×20 pixels. The pixel data within the first area 21 is obtained, and the pixel average values ​​of the RGB three channels of the pixels within the range are calculated, which is recorded as Get the maximum and minimum values ​​of the RGB channels of the pixels in the second area 22 and the third area 23, denoted as ((R 2max , R 2min ), (G 2max , G 2min ), (B 2max , B 2min )), thus obtaining a set of R2G2B2 three-channel pixel similarity ranges, namely:

[0088]

[0089]

[0090]

[0091] The RGB three-channel data of all pixels in the bird's-eye view image are obtained. According to the above pixel similarity range, the FloodFill algorithm is used to perform pixel similarity calibration on all pixels in the bird's-eye view image to form a calibration set. Then, any one set of R2G2B2 values ​​is selected within the above pixel similarity range, and the calibration set is used as a mask to cover the pixels in the calibration set with any selected set of R2G2B2 values. In this way, the lower side surface 2 in the bird's-eye view image can be identified to the greatest extent, as shown in FIG4(b). All pixels in the calibration set can be excluded to obtain the image information of the real object in the bottle holder.

[0092] Embodiment 3

[0093] In the third embodiment, the left side 3 in the bird's-eye view image of the refrigerator bottle holder is a glass surface, as shown in FIG5(a). The real object in the bottle holder has its mirror image formed on the left side 3, which interferes with the final system's acquisition of the real object information in the bottle holder. In this regard, using the above-mentioned object recognition method, in step S31, any three areas on the leftmost side of the bird's-eye view image are selected, preferably the lower left, middle left, and upper left three areas, which are recorded as the first area 31, the second area 32, and the third area 33.

[0094] Similarly, the first area 31, the second area 32, and the third area 33 are preferably in the range of 20×20 pixels. The pixel data within the first area 31 is obtained, and the pixel average value of the RGB three channels of the pixel points within the range is calculated, which is recorded as Get the maximum and minimum values ​​of the RGB channels of the pixels in the second area 32 and the third area 33, denoted as ((R 3max , R 3min ), (G 3max , G 3min ), (B 3max , B 3min )), thus obtaining a set of R3G3B3 three-channel pixel similarity ranges, namely:

[0095]

[0096]

[0097]

[0098] The RGB three-channel data of all pixels in the bird's-eye view image are obtained. According to the above pixel similarity range, the FloodFill algorithm is used to calibrate the pixel similarity of all pixels in the bird's-eye view image to form a calibration set. Then, any one group of R3G3B3 values ​​is selected within the above pixel similarity range, and the calibration set is used as a mask to cover the pixels in the calibration set with any selected group of R3G3B3 values. In this way, the left side 3 in the bird's-eye view image can be marked to the greatest extent, as shown in FIG5(b). All pixels in the calibration set can be excluded to obtain the image information of the real object in the bottle holder.

[0099] Embodiment 4

[0100] In the fourth embodiment, one side 4 on the right side in the bird's-eye view image of the bottle holder in the refrigerator is a glass surface, as shown in FIG6(a). The real object in the bottle holder has its mirror image formed on the right side 4, which interferes with the final system's acquisition of the real object information in the bottle holder. In this regard, using the above-mentioned object recognition method, in step S31, any three areas on the rightmost side of the bird's-eye view image are selected, preferably the upper rightmost, rightmost middle, and lower rightmost three areas, which are recorded as the first area 41, the second area 42, and the third area 43.

[0101] Similarly, the first area 41, the second area 42, and the third area 43 are preferably in the range of 20×20 pixels. The pixel data within the range of the first area 41 is obtained, and the pixel average value of the RGB three channels of the pixel points within the range is calculated, which is recorded as Get the maximum and minimum values ​​of the RGB channels of the pixels in the second area 42 and the third area 43, and record them as ((R 4max , R 4min ), (G 4max , G 4min ), (B 4max , B 4min )), thus obtaining a set of R4G4B4 three-channel pixel similarity ranges, namely:

[0102]

[0103]

[0104]

[0105] The RGB three-channel data of all pixels in the bird's-eye view image are obtained. According to the above pixel similarity range, the FloodFill algorithm is used to perform pixel similarity calibration on all pixels in the bird's-eye view image to form a calibration set. Then, any one set of R4G4B4 values ​​is selected within the above pixel similarity range, and the calibration set is used as a mask to cover the pixels in the calibration set with any selected set of R4G4B4 values. In this way, the right side 4 in the bird's-eye view image can be marked to the greatest extent, as shown in FIG6( b). All pixels in the calibration set can be excluded to obtain the image information of the real object in the bottle holder.

[0106] Furthermore, after excluding all pixels in the calibration set and obtaining the image information of the real object in the bottle holder, the method further comprises the steps of:

[0107] S6: Counting the number of real objects in the bird's-eye view image.

[0108] The present invention also provides a refrigerator, which comprises a bottle holder, a camera, a memory and a processor.

[0109] Furthermore, the bottle seat includes a bottle seat top, a bottle seat bottom and a bottle seat wall, and the camera is arranged in the middle of the bottle seat top, vertically downward.

[0110] The bottom of the bottle seat is used for placing articles, and at least one wall surface of the bottle seat is a glass surface.

[0111] The memory stores a computer program that can be run on the processor, and when the processor executes the program, the steps of the above-mentioned method for identifying information of items in the refrigerator are implemented.

[0112] The present invention also provides a computer storage medium, in which a computer program is stored, and when the computer program is run, the steps of the above-mentioned method for identifying information of items in a refrigerator are executed on the device where the computer storage medium is located.

[0113] In summary, the present invention improves the brightness of the bird's-eye view of the space inside the bottle holder, calculates the pixel similarity range based on the pixel value data of the outer area, and then uses the FloodFill algorithm to calibrate the pixel similarity of all pixels in the image to form a calibration set. The pixel point data in the calibration set can be excluded, that is, the mirror area of ​​the bird's-eye view of the refrigerator bottle holder is removed, thereby eliminating the interference of the mirrored objects, obtaining the image information of the real objects in the bottle holder, and improving the accuracy of the recognition algorithm.

[0114] It should be understood that although this specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each implementation mode may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

[0115] The series of detailed descriptions listed above are only specific descriptions of feasible implementation methods of the present invention. They are not intended to limit the scope of protection of the present invention. Any equivalent implementation methods or changes that do not deviate from the technical spirit of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for identifying information of items in a refrigerator, characterized in that: Includes steps: Get a bird's-eye view of the space inside the refrigerator bottle holder; Increase the brightness of the bird's-eye view image to 1.5 to 2.5 times of the original brightness; Selecting an outer area of ​​the bird's-eye view image, and determining the pixel similarity range according to pixel value data in the area; Performing pixel similarity calibration on all pixels in the bird's-eye view image according to the pixel similarity range to form a calibration set, wherein the calibration set includes all pixel data in the bird's-eye view image that fall within the pixel similarity range; All pixel points in the calibration set are excluded to obtain image information of the real object in the bottle holder.

2. The method for identifying information of items in a refrigerator according to claim 1, characterized in that: "Obtaining a bird's-eye view of the space inside the refrigerator bottle holder" specifically includes: A camera arranged in the middle of the top of the bottle seat is controlled to capture a bird's-eye view image of the space inside the bottle seat.

3. The method for identifying information of items in a refrigerator according to claim 2, characterized in that: “Selecting the outer area of ​​the bird's-eye view image” specifically includes: Select any three outermost areas of the bird's-eye view image, where the three areas are spaced a certain distance apart; The three areas include a first area, a second area and a third area, wherein: The first area and the third area are located at two outermost ends of the bird's-eye view image, respectively, and the second area is located at the outermost middle position of the bird's-eye view image.

4. The method for identifying information of items in a refrigerator according to claim 3, characterized in that: “Determining the pixel similarity range according to the pixel value data in the region” specifically includes the steps of: Obtain pixel values ​​within any area of ​​the three areas, and calculate an average value of the pixel values; Obtain pixel values ​​within the other two regions, and calculate the maximum and minimum values ​​of the pixels within the two regions; The pixel similarity range is determined according to the average value of pixels in the one region and the maximum and minimum values ​​of pixels in the other two regions.

5. The method for identifying information of items in a refrigerator according to claim 4, characterized in that: “Performing pixel similarity calibration on all pixels in the bird's-eye view image according to the pixel similarity range to form a calibration set” specifically includes the following steps: Obtaining pixel value data of all pixels in the bird's-eye view image; According to the pixel similarity range, a FloodFill algorithm is used to perform pixel similarity calibration on all pixels in the bird's-eye view image; Form a calibration set.

6. The method for identifying information of items in a refrigerator according to claim 5, characterized in that: “Excluding all pixels in the calibration set” specifically includes: Obtain any pixel value within the pixel similarity range, covering the calibration set.

7. The method for identifying information of items in a refrigerator according to claim 1, characterized in that: The pixel value data specifically includes R, G, and B three-channel pixel value data.

8. The method for identifying information of items in a refrigerator according to claim 1, characterized in that: The method further comprises the steps of: The number of real objects in the bird's-eye view image is counted.

9. A refrigerator, comprising: Bottle holder, camera, memory and processor, characterized in that, The bottle seat comprises a top, a bottom and a wall of the bottle seat, and the camera is arranged in the middle of the top of the bottle seat, and is arranged vertically downward; The bottom of the bottle seat is used for placing items; At least one side of the bottle seat wall is a glass surface; The memory stores a computer program that can be run on the processor, and when the processor executes the program, the steps of the method for identifying information of items in a refrigerator as described in any one of claims 1 to 8 are implemented.

10. A computer storage medium storing a computer program, characterized in that: When the computer program is executed, the device where the computer storage medium is located causes the device to execute the steps of the method for identifying information of items in a refrigerator according to any one of claims 1-8.

Images