Search

CN-115457524-B - Article identification method for refrigeration equipment, storage medium, system and refrigeration equipment

CN115457524BCN 115457524 BCN115457524 BCN 115457524BCN-115457524-B

Abstract

The invention provides an article identification method for refrigeration equipment, a storage medium, a system and the refrigeration equipment. The identification method comprises the steps of obtaining a target image of an article stored in refrigeration equipment, obtaining target features according to the target image, matching the target features with the article type features in an article type feature library in a refrigeration equipment database one by one and calculating similarity, obtaining article type naming information of the article if the similarity is lower than a first similarity value, and building a mapping relation between the article type naming information and the target features and storing the article type naming information into the article type feature library. Wherein the storage medium, the system and the refrigeration equipment can realize the identification method. By the arrangement, the refrigerating equipment after being put into use can identify new kinds of articles, and the article identification range of the refrigerating equipment is widened.

Inventors

  • BI YANHUA
  • KONG LINGLEI
  • GAO HONGBO

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
20220815

Claims (14)

  1. 1. An article identification method for a refrigeration appliance, the identification method comprising: Acquiring a target image of an article stored in refrigeration equipment; acquiring target characteristics according to the target image; matching the target features with the object type features in the object type feature library in the refrigeration equipment database one by one and calculating the similarity; if the similarity is lower than a first similarity value, acquiring article type naming information of the article, establishing a mapping relation between the article type naming information and the target feature, and storing the mapping relation into the article type feature library; if the similarity is greater than or equal to the first similarity value, acquiring the maximum similarity; Judging whether the maximum similarity is larger than a third similarity value or not, wherein the third similarity value is larger than the first similarity value; if yes, deleting the target characteristic data; if not, acquiring the article type of the article type feature corresponding to the maximum similarity, establishing a mapping relation between the article type and the target feature, and storing the mapping relation into the article type feature library; the article type feature library comprises sub-article type feature libraries which are in one-to-one correspondence with the article types, the sub-article type feature library comprises article type features corresponding to the article types, The step of acquiring the article type to which the article type feature corresponding to the maximum similarity belongs, establishing a mapping relation between the article type and the target feature, and storing the mapping relation in the article type feature library specifically comprises the following steps: And acquiring the article type to which the article type feature corresponding to the maximum similarity belongs, establishing a mapping relation between the article type and the target feature, and storing the target feature into a sub-article type feature library corresponding to the article type.
  2. 2. The article identification method for a refrigeration appliance according to claim 1, further comprising: Matching the target features with the features in the unmanaged feature library one by one and calculating the similarity; And if the similarity is larger than a second similarity value, storing the target feature into the unmanaged feature library, wherein the second similarity value is larger than or equal to the first similarity value.
  3. 3. The article identification method for a refrigeration apparatus according to claim 2, wherein "acquiring article type naming information of the article" includes: And outputting the item type naming prompt information and receiving the input item type naming information.
  4. 4. The article identification method for a refrigeration appliance of claim 3, further comprising: and if the naming information is not received, storing the target feature into the unmanaged feature library.
  5. 5. The article identification method for a refrigeration appliance according to claim 1, further comprising: Acquiring the object type features corresponding to the maximum similarity as historical maximum similarity, and storing the total number of historical target features of the sub object type feature library; and if the total number reaches a preset value, deleting the target feature.
  6. 6. The article identification method for a refrigeration appliance of claim 5, further comprising: acquiring a similarity value interval in which the maximum similarity is located; obtaining the total number of the historical maximum similarity corresponding to the similarity value interval; And deleting the target feature if the total number reaches a section set value, wherein the section set value is inversely proportional to the size of the similarity value section.
  7. 7. The article identification method for a refrigeration appliance according to claim 1, further comprising: and acquiring the total number of the article type features in the sub-article type feature library, and deleting the target features if the total number reaches a preset value.
  8. 8. The article identification method for a refrigeration appliance according to claim 1, further comprising: if the similarity is greater than or equal to the first similarity value, acquiring the maximum similarity; judging whether the maximum similarity is lower than a fourth similarity value or not, wherein the fourth similarity value is larger than the first similarity value; If so, acquiring the article type of the article type feature corresponding to the maximum similarity, and outputting prompt information for confirming whether the target article belongs to the article type; and if the instruction information that the target does not belong to the article type is acquired, acquiring article type naming information of the article, establishing a mapping relation between the target feature and the article type naming information, and storing the mapping relation into the article type feature library.
  9. 9. The method for item identification of refrigeration equipment of claim 1, further comprising obtaining a cumulative unmatched number of days for each item type in said refrigeration equipment item type database; if the accumulated unmatched days reach a first preset day value, outputting prompt information of whether the article type needs to be managed; If the instruction information of the article type to be managed is acquired, reserving the data of the article type to a second preset day value; if instruction information is acquired that does not require management of the article type, the article type characteristic data of the article type is deleted.
  10. 10. The article identification method for a refrigeration appliance according to claim 1, further comprising: the target feature and the object type feature are N-dimensional feature vectors with the same dimension, and the N dimensions comprise 512 dimensions or 256 dimensions.
  11. 11. The article identification method for a refrigeration apparatus according to claim 1, wherein "acquiring a target image of an article stored in said refrigeration apparatus" specifically includes: acquiring an image shot by refrigeration equipment; inputting the image into a detection model, and unifying the inherent structure of the refrigeration equipment in the image as a prospect; And acquiring a target image of the object stored in the refrigeration equipment from the foreground.
  12. 12. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, carries out the steps in the item identification method for a refrigeration appliance according to any one of claims 1 to 11.
  13. 13. An article identification system comprising a control module, an article identification module, characterized in that the article identification system further comprises a memory and a processor, the memory storing a computer program executable on the processor, the processor implementing the steps in the article identification method for a refrigeration appliance according to any of claims 1-11 when the computer program is executed.
  14. 14. A refrigeration appliance comprising an item identification system comprising a control module, an item identification module, characterized in that the item identification system further comprises a memory and a processor, the memory storing a computer program executable on the processor, the processor implementing the steps in the item identification method for a refrigeration appliance according to any of claims 1-11 when the computer program is executed.

Description

Article identification method for refrigeration equipment, storage medium, system and refrigeration equipment Technical Field The present invention relates to the field of household appliances, and in particular, to an article identification method for a refrigeration device, a storage medium, a system, and a refrigeration device. Background Along with the progress of science and technology, the requirements of users on refrigeration equipment are higher and higher, and intelligent transformation becomes a new research and development direction of the refrigeration equipment. Refrigeration equipment, such as refrigerators, is typically provided with an item identification system. The article identification system generally comprises a camera arranged in the refrigeration equipment, and the articles stored in the refrigeration equipment are identified by identifying the pictures of the articles in the storage room taken by the camera so as to intelligently manage the articles in the refrigeration equipment. In the article identification system of the refrigeration equipment in the prior art, identification training of a specific type of article is usually required to be performed in advance before delivery, so that the refrigeration equipment can identify the type of article. However, such a design has a disadvantage in that articles which may be stored during the use of the refrigerating apparatus by the user are various, such as new products, etc., and thus it is impossible to secure the coverage of all articles in the article type recognition training before shipment, resulting in the fact that the refrigerating apparatus cannot recognize and manage some types of articles. Disclosure of Invention The invention aims to provide an article identification method, a storage medium, a system and a refrigeration device for the refrigeration device, wherein the article identification method, the storage medium, the system and the refrigeration device can realize that the refrigeration device after being put into use can identify new types of articles by acquiring article type naming information of the articles, establishing a mapping relation between the article type naming information and target characteristics of the articles and storing the article type naming information into an article type characteristic library, and expand the article identification range of the refrigeration device. To achieve the above object, an embodiment of the present invention provides an article identification method for a refrigeration apparatus, wherein the identification method includes: Acquiring a target image of an article stored in refrigeration equipment; acquiring target characteristics according to the target image; matching the target features with the object type features in the object type feature library in the refrigeration equipment database one by one and calculating the similarity; and if the similarity is lower than a first similarity value, acquiring the article type naming information of the article, establishing a mapping relation between the article type naming information and the target feature, and storing the mapping relation into the article type feature library. As a further improvement of an embodiment of the present invention, the method further comprises: Matching the target features with the features in the unmanaged feature library one by one and calculating the similarity; And if the similarity is larger than a second similarity value, storing the target feature into the unmanaged feature library, wherein the second similarity value is larger than or equal to the first similarity value. As a further improvement of an embodiment of the present invention, wherein "acquiring item type naming information of the item" includes: And outputting the item type naming prompt information and receiving the input item type naming information. As a further improvement of an embodiment of the present invention, the method further comprises: and if the naming information is not received, storing the target feature into the unmanaged feature library. As a further improvement of an embodiment of the present invention, the method further comprises: if the similarity is greater than or equal to the first similarity value, acquiring the maximum similarity; Judging whether the maximum similarity is larger than a third similarity value or not, wherein the third similarity value is larger than the first similarity value; if yes, deleting the target characteristic data; if not, acquiring the article type corresponding to the maximum similarity, establishing a mapping relation between the article type and the target feature, and storing the mapping relation in the article type feature library. As a further improvement of an embodiment of the present invention, the article type feature library includes sub-article type feature libraries corresponding to the article types one by one, the sub-article type feature libraries inc