Search

CN-113255428-B - Article discriminating system

CN113255428BCN 113255428 BCN113255428 BCN 113255428BCN-113255428-B

Abstract

The article discrimination system includes an identification information acquisition unit, an imaging unit that images the appearance of the object article, a determination unit that derives the degree of coincidence between the imaged image of the object article and the appearance image data stored in the storage unit and determines whether or not the appearance image data has a degree of coincidence equal to or greater than a determination threshold, and a learning processing unit that performs learning processing when the determination unit determines that the appearance image data has not been in the normal state. The learning process is a process of storing the captured image data in the storage unit as new appearance image data in association with the identification information acquired by the identification information acquisition unit. The storage unit stores a predetermined number of appearance image data of 2 or more in association with 1 piece of identification information in the order from new to old at the time points stored in the storage unit.

Inventors

  • Qing Chuanshe
  • Ko keigo
  • Wei Shengchun
  • ITO KAZUTOSHI

Assignees

  • 株式会社大福

Dates

Publication Date
20260505
Application Date
20210210
Priority Date
20200212

Claims (6)

  1. 1. An article discrimination system, characterized in that: The article discrimination system includes: A storage unit that stores appearance image data, which is data of an appearance image of an article, in association with identification information of the article; An identification information acquisition unit that acquires identification information of a target object, which is an object to be discriminated; A photographing unit that photographs an appearance of the target object; A determination unit configured to derive, as the appearance image data stored in the storage unit, a degree of coincidence between the appearance image data associated with the identification information of the object acquired by the identification information acquisition unit and the captured image of the object captured by the capturing unit, and determine whether or not the appearance image data having the degree of coincidence equal to or greater than a determination threshold is in a normal state, and A learning processing unit that executes learning processing when the determination unit determines that the normal state is not the normal state; The learning process is a process of storing, in the storage unit, new appearance image data of the captured image of the object article in which the appearance image data having the degree of coincidence equal to or higher than the determination threshold is not present, in association with the identification information of the object article acquired by the identification information acquisition unit, The storage unit stores 2 or more pieces of predetermined number of pieces of appearance image data in association with 1 piece of identification information in order from new to old at the time points stored in the storage unit.
  2. 2. The article discrimination system of claim 1, wherein, The determination unit executes a matching degree derivation process of deriving the matching degree in order of the appearance image data from new to old at the time points stored in the storage unit, and ends the matching degree derivation process when the appearance image data whose matching degree is equal to or higher than the determination threshold is detected, and determines that the appearance image data is in the normal state.
  3. 3. The article discrimination system of claim 1, wherein, Further comprising a dimension measuring unit for measuring the dimension of the object article, Size data, which is data of the size of an article that may become the target article, is collected in advance and stored in the storage section in association with identification information of the article, The learning processing unit executes the learning processing on the condition that the size of the object measured by the size measuring unit is stored in an allowable size range when the determination unit determines that the object is not in the normal state, and does not execute the learning processing on the condition that the size of the object measured by the size measuring unit is not stored in the allowable size range, the allowable size range being based on the size data stored in the storage unit in association with the identification information acquired by the identification information acquiring unit.
  4. 4. The article discrimination system of claim 3 wherein, Further comprising an operation unit for moving the object, The object article is an article which is included in a group of a plurality of articles of the same kind placed on the support body and on which other articles are not placed, The size of the object article is measured by the size measuring section in a state where the object article is moved by the operating section so that a gap is formed between the object article and the other adjacent article.
  5. 5. The article discrimination system of claim 3 wherein, The dimension measuring unit is provided with a TOF (Time Of Flight) camera or a stereo camera.
  6. 6. The article discrimination system according to any one of claims 1 to 5, wherein, The object article is an article which is included in an article group which is a group of a plurality of articles placed on a support body and on which other articles are not placed, Further comprising a removal processing unit that performs a removal process of sequentially selecting the target articles from the article group and removing the selected target articles from the support, The determination unit determines whether or not the object is in the normal state based on the photographed image of the newly selected object every time the object to be the object of the extraction processing is selected by the extraction processing unit, The extraction processing unit extracts the selected object from the support body when the determination unit determines that the object is in the normal state or when the learning processing unit performs the learning processing when the determination unit determines that the object is not in the normal state.

Description

Article discriminating system Technical Field The present invention relates to an article discrimination system using appearance image data of an article. Background An example of the above-described article discrimination system is disclosed in japanese patent application laid-open No. 2015-43175 (patent document 1). In the following description of the related art, symbols shown in brackets are symbols of patent document 1. In patent document 1, an article discriminating system is applied to an article discriminating device (3) for discriminating the position of an article (50) to be extracted from a scene where the article (50) is extracted from a container (51). The article identification device (3) identifies the position of the article (50) using a template image of a pattern including the outer surface of the article (50). Specifically, the article identification device (3) detects, as an article-existing region (M), a region included in a captured image of a container (51) containing an article (50) and having a degree of coincidence with a template image of a predetermined threshold value or more, thereby identifying the position of the article (50). As described above, patent document 1 discloses an article discrimination system using a template image as appearance image data. Disclosure of Invention However, there are cases where the design of the outer surface of the article such as the pattern is changed even for the same kind of article. For example, in the case where the article is a container for containing a commodity (for example, a cardboard box for containing a food product), even if the article is a container of the same type (specifically, a container for containing a commodity of the same type), the design (package design) of the outer surface of the container may be changed according to a season, sales promotion, or the like. In order to perform appropriate discrimination in the article discrimination system, when the design of the article is changed, appearance image data corresponding to the changed design is required. In this case, however, since the frequency of generation and registration of the appearance image data is high, there is a possibility that the labor and time of an operator who operates the article discrimination system are increased, the processing speed of an apparatus to which the article discrimination system is applied (in patent document 1, the article identification apparatus) is reduced, and the like. Accordingly, it is desirable to realize an article discrimination system capable of suppressing the frequency of generation and registration of the appearance image data to be low. The article discrimination system of the present disclosure includes a storage unit that stores appearance image data, which is data of an appearance image of an article, in association with identification information of the article, an identification information acquisition unit that acquires identification information of an article to be discriminated, an imaging unit that images an appearance of the article to be discriminated, a determination unit that derives a degree of coincidence between an imaged image of the article to be imaged by the imaging unit and the appearance image data stored in the storage unit in association with the identification information acquired by the identification information acquisition unit, determines whether or not there is a normal state of the appearance image data in which the degree of coincidence is equal to or greater than a determination threshold, and a learning processing unit that performs a learning process of associating 1 identification information with 2 or more pieces of old appearance image data in order from new to old in time point stored in the storage unit, by associating the data of the imaged image as new appearance image data with the identification information acquired by the identification information acquisition unit. According to this configuration, when the determination unit determines that the image is not in the normal state, that is, when the appearance image data having a degree of coincidence with the captured image of the target object equal to or greater than the determination threshold value does not exist in the storage unit, the captured image of the target object can be stored in the storage unit as new appearance image data in association with the identification information of the target object by the learning process. Therefore, when the design of the object is changed, the design image data corresponding to the changed design can be generated and registered. This can reduce the labor and time of the worker compared with the case where the worker manually performs the job of generating and registering the appearance image data corresponding to the changed appearance design. However, there are cases where the design of the object is periodically changed between a plurality of designs or the object is temporari