CN-115200311-B - Method for identifying article information in refrigerator, refrigerator and computer storage medium

CN115200311BCN 115200311 BCN115200311 BCN 115200311BCN-115200311-B

Abstract

The invention discloses a method for identifying information of articles in a refrigerator, the refrigerator and a computer storage medium, wherein the method comprises the steps of obtaining an aerial view image of a space in a bottle seat of the refrigerator; the method comprises the steps of locating character areas on the upper surfaces of objects in the aerial view image, constructing identification frames, wherein the identification frames correspond to the character areas on the upper surfaces of objects, each identification frame comprises all character information in the corresponding character area, extracting the identification frames, keeping the shapes of the identification frames unchanged, splicing the identification frames, constructing areas to be identified, comprising all the identification frames, and identifying the character information in the areas to be identified. The invention can output a plurality of character information through one-time recognition algorithm, and improves the efficiency of the recognition algorithm.

Inventors

GAO HONGBO
KONG LINGLEI

Assignees

青岛海尔电冰箱有限公司
海尔智家股份有限公司

Dates

Publication Date: 20260512
Application Date: 20220627

Claims (7)

1. The method for identifying the information of the articles in the refrigerator is characterized by comprising the following steps: Acquiring a bird's eye view image of the space in the bottle seat of the refrigerator; Positioning a character area on the upper surface of each article in the aerial view image; Constructing identification frames, wherein the identification frames correspond to the character areas on the upper surface of an article, and each identification frame contains all character information in the corresponding character area; Extracting the identification frames, keeping the shape of the identification frames unchanged, keeping the relative position relation of each identification frame in the aerial view image unchanged, fixing the position of one identification frame unchanged, sequentially moving all other identification frames towards the direction of the identification frame to perform splicing treatment, stopping moving when the identification frame moves to be contacted with the other identification frame, and splicing each identification frame; Constructing a region to be identified containing all the identification frames; identifying character information in the area to be identified; the step of extracting the identification frames, keeping the shapes of the identification frames unchanged, and splicing the identification frames specifically comprises the following steps: carrying out tight splicing treatment on all the identification frames in sequence along the transverse direction and in sequence along the longitudinal direction; the method further comprises the steps of: the recognition results are one-to-one corresponding to each recognition frame, and character information in each recognition frame is obtained; Corresponding to a character area on the upper surface of each article in the aerial view image, and obtaining character information on the upper surface of each article in the aerial view image; The character information corresponding to the character area of the upper surface of the object in the aerial view image is obtained, and the character information of the upper surface of each object in the aerial view image is obtained specifically comprises the following steps: detecting the content of the character information about date information, and corresponding the date information to each identification frame to obtain the date information in each identification frame; And obtaining date information of the upper surfaces of the articles in the aerial view image according to the character areas of the upper surfaces of the articles in the aerial view image corresponding to the identification frames.
2. The method for identifying information of articles in a refrigerator according to claim 1, wherein said positioning the character area on the upper surface of the articles in the bird's eye view image specifically comprises: and positioning a character area on the upper surface of the object in the aerial view image by using a character positioning algorithm.
3. The method for identifying information of articles in a refrigerator according to any one of claims 1 to 2, wherein the construction of the area to be identified including the identification frame specifically includes: constructing the area to be identified in a rectangular structure; And taking the maximum value of the overall length of all the spliced identification frames in the horizontal direction as the length of the area to be identified, and taking the maximum value of the overall length of all the spliced identification frames in the vertical direction as the width of the area to be identified.
4. The method for identifying information of articles in a refrigerator according to claim 3, wherein said identifying character information of said area to be identified specifically comprises: and identifying all character information in the area to be identified by utilizing an OCR (optical character recognition) algorithm.
5. The method for recognizing article information in a refrigerator according to claim 1, wherein the article information includes article type information; the method further comprises the steps of: A target detection algorithm is utilized to mark a detection frame corresponding to each article in the aerial view image; acquiring the positions of the articles in the detection frames in the refrigerator bottle seats according to the calibration positions of the detection frames; And detecting and identifying the outline type of the article in each detection frame, and identifying the article category information.
6. A refrigerator comprises a bottle seat, a camera, a memory and a processor, and is characterized in that, The camera is arranged in the middle of the top of the bottle seat and is vertically arranged downwards; The bottom of the bottle seat is used for placing articles; the memory stores a computer program executable on the processor, which when executed, implements the steps of the method for identifying information of articles in a refrigerator as claimed in any one of claims 1 to 5.
7. A computer storage medium in which a computer program is stored, characterized in that the computer program, when run, causes a device in which the computer storage medium is located to perform the steps of the method for identifying information of articles in a refrigerator according to any one of claims 1-5.

Description

Method for identifying article information in refrigerator, refrigerator and computer storage medium Technical Field The present invention relates to the field of refrigeration apparatuses, and in particular, to a method for identifying information of articles in a refrigerator, and a computer storage medium. Background With the development of electrical appliances intellectualization, the identification of internal articles has been a necessary function of intelligent refrigerators. In order to realize identification of articles stored in the refrigerator, one or more cameras are arranged in the refrigerator and used for collecting photos of the articles in the refrigerator, and the collected photos are sent to a user terminal after identification processing so as to be checked by a user. The camera is arranged above the articles, so that the problem that the articles are mutually shielded can be avoided, and the accuracy of article information identification is improved. There are a number of problems in the implementation of intelligent refrigerators, for example, in terms of how to identify the text information of the items in the refrigerator bottle holder, including the item name, date of manufacture, date of expiration, etc. The existing method is that the shot space image in the bottle seat of the refrigerator is wholly substituted into an algorithm to locate the specific position of the article text information on the image, and then the article text information is identified, and then the next article text information is identified. When the number of articles in the bottle seat of the refrigerator is large, the time consumption of the method for identifying the articles one by one is multiplied, and the identification efficiency is low. Disclosure of Invention The invention aims to provide a method for identifying information of articles in a refrigerator, the refrigerator and a computer storage medium. In order to achieve one of the above objects, an embodiment of the present invention provides a method for identifying information of articles in a refrigerator, including the steps of: Acquiring a bird's eye view image of the space in the bottle seat of the refrigerator; Positioning a character area on the upper surface of each article in the aerial view image; Constructing identification frames, wherein the identification frames correspond to the character areas of the surface of an article, and each identification frame contains all character information in the corresponding character area; extracting the identification frames, keeping the shape of the identification frames unchanged, and splicing the identification frames; Constructing a region to be identified containing all the identification frames; and identifying character information in the area to be identified. As a further improvement of an embodiment of the present invention, the positioning the character area on the upper surface of the object in the bird's eye view image specifically includes: and positioning a character area on the upper surface of the object in the aerial view image by using a character positioning algorithm. As a further improvement of an embodiment of the present invention, the extracting the identification frames, keeping the shape of the identification frames unchanged, and splicing the identification frames specifically includes: maintaining the relative position relation of each identification frame in the aerial view image unchanged, and extracting each identification frame; fixing the position of one identification frame without changing, sequentially moving all other identification frames towards the direction of the identification frame for splicing, and stopping moving when the identification frame moves to be contacted with the other identification frame. As a further improvement of an embodiment of the present invention, the extracting the identification frames, keeping the shape of the identification frames unchanged, and splicing the identification frames specifically includes: and (5) carrying out tight splicing treatment on all the identification frames in sequence along the transverse direction and in sequence along the longitudinal direction. As a further improvement of an embodiment of the present invention, the constructing the area to be identified including the identification frame specifically includes: constructing the area to be identified in a rectangular structure; And taking the maximum value of the overall length of all the spliced identification frames in the horizontal direction as the length of the area to be identified, and taking the maximum value of the overall length of all the spliced identification frames in the vertical direction as the width of the area to be identified. As a further improvement of an embodiment of the present invention, the identifying the character information of the area to be identified specifically includes: and identifying all character information i