Search

CN-116863155-B - Method, device, equipment and storage medium for identifying articles in bin

CN116863155BCN 116863155 BCN116863155 BCN 116863155BCN-116863155-B

Abstract

The application discloses a method, a device, equipment and a storage medium for identifying articles in a bin. The method comprises the steps of obtaining a plurality of first identification image information of articles in a bin and first identification data information of the articles collected by an RFID sensor in the bin, determining article identification data according to the first identification image information and the first identification data information, determining first article identification data according to the first identification data information, determining second article identification data according to the first identification image information, and taking intersection of the first article identification data and the second article identification data as article identification data. The application can improve the efficiency of identifying the articles in the bin, can achieve the aims of accurate identification and quick identification, and can effectively prevent the problem of false identification of the articles.

Inventors

  • SHI BENCAI

Assignees

  • 北京应天海乐科技发展有限公司

Dates

Publication Date
20260508
Application Date
20230608

Claims (8)

  1. 1. A method of identifying items in a bin, the method comprising: acquiring a plurality of first identification image information of articles in a bin and first identification data information of the articles acquired by an RFID sensor in the bin, wherein the first identification image information is the image information of the articles in the bin acquired from a plurality of angles through an image acquisition device; determining article identification data according to the first identification image information and the first identification data information; Wherein, determining the article identification data according to the first identification image information and the first identification data information includes: determining first article identification data from the first identification data information; determining second article identification data from the first identification image information; taking intersection of the first article identification data and the second article identification data as article identification data; The determining second article identification data from the first identification image information includes: Performing first processing on the first identification image information to obtain a first processed image, wherein the first processing is to perform correction processing on the first identification image information; Performing second processing on the first processed image to obtain a second processed image, wherein the second processing is to perform image enhancement processing on the first processed image; Comparing the second processed image with image sample data in a picture library, setting a similarity threshold of the second processed image and the image sample data, and obtaining second object identification data; The first processing is performed on the first identification image information to obtain a first processed image, and the first processing is performed on the first identification image information, and includes: correcting adjacent images in the first identification image information, wherein the adjacent images are two adjacent images acquired by the same image acquisition equipment, and the step of correcting the adjacent images comprises the following steps of: any frame of image in the first identification image information is obtained as a basic image; calculating the similarity value of a central image block in the basic image and any image block in an adjacent image of the basic image: ; Wherein, the Is the pixel mean value of the kth image block in the adjacent image of the base image, Is the pixel mean of the base image block, A pixel mean value of any image block of adjacent images of the base image; Similarity value Highest corresponding image block As a central image block Matching image blocks in adjacent images; Respectively extracting And SIFT feature points in 8 directions are matched, and a { set of the feature points is obtained And } wherein, For a set of matched SIFT feature points, As a base image At the SIFT feature point in the direction 1, Adjacent images as base images Intermediate and base images A kind of electronic device And constructing a feature point matching matrix by matching the feature points: ; according to the relation between the characteristic point matching matrix and the basic processing matrix F, singular value decomposition is carried out on the characteristic point matching matrix, and a singular matrix corresponding to the minimum singular value after decomposition is the solution of the basic processing matrix F; From the solution of the basic processing matrix F, we get A pole of (2); performing matrix transformation on adjacent images of the base image, and simultaneously converting parallax of the base image into a depth value; and taking the converted image as a corrected imaging result, and carrying out rectangular transformation correction on the imaging result of the next adjacent image by using the corrected basic image to obtain a first processed image.
  2. 2. The method according to claim 1, wherein the method further comprises: Acquiring quantity information and quality information of articles in a bin; Acquiring third article identification data according to the quantity information and the quality information of the articles; And determining the article identification data according to the first article identification data, the second article identification data and the third article identification data.
  3. 3. The method of claim 2, wherein the determining item identification data based on the first item identification data, second item identification data, and third item identification data comprises: intersection of the first article identification data and third article identification data; and acquiring the intersection with the second article identification data to obtain article identification data.
  4. 4. The method of claim 1, wherein the comparing the second processed image with image sample data in a picture library, setting a similarity threshold for the second processed image and the image sample data, and obtaining second article identification data, comprises: Respectively comparing the image characteristic information of the second processed image with the image characteristic information of each image sample data in the image library, selecting a first target image sample with similarity meeting a first setting condition from the image sample data in the image library, and selecting a second target image sample with similarity meeting a second setting condition from the image sample data in the image library; Determining an image structural feature difference of an average image structural feature of the second target image sample relative to an average image structural feature of the first target image sample, and determining an image colorimetric feature difference of an average image colorimetric feature of the second target image sample relative to an average image colorimetric feature of the first target image sample; Performing image structure processing and image chromaticity processing on the second processed image according to the image structure feature differences and the image chromaticity feature differences to respectively obtain a third processed image and a fourth processed image of the second processed image which are matched with the first target image sample and the second target image sample; and respectively comparing the similarity of the third processed image and the first target sample image, and the similarity of the fourth processed image and the second target sample image, and acquiring second object identification data according to the set similarity threshold.
  5. 5. The method of claim 4, wherein selecting a first target image sample from the image sample data in the picture library having a similarity satisfying a first set condition, and selecting a second target image sample from the image sample data in the picture library having a similarity satisfying a second set condition, comprises: Comparing the description features of the second processed image with the description features of each sample, screening candidate first target sample images with the similarity meeting a first sub-condition from image sample data in the image library according to the sequence of the similarity from high to low, and screening candidate second target image samples with the similarity meeting a second sub-condition from the image sample data in the image library, wherein the first sub-condition is a front view of the image, and the second sub-condition is a top view of the image; And matching the key point information of the second processed image with the similarity of the key points of each image sample in the candidate first target image sample and the candidate second target image sample respectively, screening the first target image with the similarity of the key points meeting a set threshold value from the candidate first target image sample according to the sequence of the similarity of the key points from high to low, and screening the second target image sample with the similarity of the key points meeting the set threshold value from the candidate second target image sample.
  6. 6. The method of claim 1, wherein performing a second process on the first processed image to obtain a second processed image, the second process being an image enhancement process on the first processed image, comprises: performing gamma enhancement processing on each of the first processed images, the gamma enhancement processing being: ; in the formula, For one of the second processed images, The gamma parameter is the image sequence of the second processed image, at this time, is: 。
  7. 7. an in-bin item identification device, the device comprising: The acquisition module is used for acquiring a plurality of first identification image information of the articles in the bin and first identification data information of the articles acquired by the RFID sensor in the bin, wherein the first identification image information is the image information of the articles in the bin acquired from a plurality of angles through the image acquisition equipment; The identification module is used for determining article identification data according to the first identification image information and the first identification data information, wherein the step of determining the article identification data according to the first identification image information and the first identification data information comprises the steps of determining first article identification data according to the first identification data information, determining second article identification data according to the first identification image information, and taking an intersection of the first article identification data and the second article identification data as article identification data; The first processing is used for carrying out first processing on the first identification image information to obtain a first processed image, and the first processing is correction processing on the first identification image information; Performing second processing on the first processed image to obtain a second processed image, wherein the second processing is to perform image enhancement processing on the first processed image; Comparing the second processed image with image sample data in a picture library, setting a similarity threshold of the second processed image and the image sample data, and obtaining second object identification data; The step for correcting the adjacent images in the first identification image information, wherein the adjacent images are two adjacent images acquired by the same image acquisition equipment, and the step for correcting the adjacent images comprises the following steps: any frame of image in the first identification image information is obtained as a basic image; calculating the similarity value of a central image block in the basic image and any image block in an adjacent image of the basic image: ; Wherein, the Is the pixel mean value of the kth image block in the adjacent image of the base image, Is the pixel mean of the base image block, A pixel mean value of any image block of adjacent images of the base image; Similarity value Highest corresponding image block As a central image block Matching image blocks in adjacent images; Respectively extracting And SIFT feature points in 8 directions are matched, and a { set of the feature points is obtained And } wherein, For a set of matched SIFT feature points, As a base image At the SIFT feature point in the direction 1, Adjacent images as base images Intermediate and base images A kind of electronic device And constructing a feature point matching matrix by matching the feature points: ; according to the relation between the characteristic point matching matrix and the basic processing matrix F, singular value decomposition is carried out on the characteristic point matching matrix, and a singular matrix corresponding to the minimum singular value after decomposition is the solution of the basic processing matrix F; From the solution of the basic processing matrix F, we get A pole of (2); performing matrix transformation on adjacent images of the base image, and simultaneously converting parallax of the base image into a depth value; and taking the converted image as a corrected imaging result, and carrying out rectangular transformation correction on the imaging result of the next adjacent image by using the corrected basic image to obtain a first processed image.
  8. 8. An electronic device, characterized in that, the device includes one or more processors and memory, the memory is used for storing one or more programs; the one or more programs, when executed by the processor, cause the processor to implement the method of any of claims 1-6.

Description

Method, device, equipment and storage medium for identifying articles in bin Technical Field The present disclosure relates generally to the field of image recognition technology, and in particular, to a method, an apparatus, a device, and a storage medium for recognizing articles in a bin. Background With the improvement of the automation technology of medical equipment, the automatic storage, identification and distribution of some medical equipment are possible, and in the prior art, the storage, identification and distribution of the medical equipment are mainly performed manually, for example, two-dimensional codes and RFID labels are arranged on the medical equipment, or the visual identification technology of a convolutional neural network is adopted, or some sensor equipment is adopted, so that the methods are complex, effective management of medical materials is difficult to realize, and particularly, the materials in a bin are accurately identified for the storage bin of some medical materials, so that the material management efficiency can be improved. Disclosure of Invention In view of the foregoing drawbacks or shortcomings of the prior art, it is desirable to provide a method, apparatus, device and storage medium for identifying articles in a bin, which meet the needs of the art. According to an aspect of the embodiment of the application, the embodiment of the application provides a method for identifying articles in a bin, which comprises the following steps: acquiring a plurality of first identification image information of articles in a bin and first identification data information of the articles acquired by an RFID sensor in the bin, wherein the first identification image information is the image information of the articles in the bin acquired from a plurality of angles through an image acquisition device; determining article identification data according to the first identification image information and the first identification data information; Wherein, determining the article identification data according to the first identification image information and the first identification data information includes: determining first article identification data from the first identification data information; determining second article identification data from the first identification image information; the intersection of the first item identification data and the second item identification data is taken as item identification data. In another embodiment of the present application, the method further comprises: Acquiring quantity information and quality information of articles in a bin; Acquiring third article identification data according to the quantity information and the quality information of the articles; And determining the article identification data according to the first article identification data, the second article identification data and the third article identification data. In another embodiment of the present application, the determining the article identification data according to the first article identification data, the second article identification data, and the third article identification data includes: intersection of the first article identification data and third article identification data; and acquiring the intersection with the second article identification data to obtain article identification data. In another embodiment of the present application, the determining the second article identification data by the first identification image information includes: Performing first processing on the first identification image information to obtain a first processed image, wherein the first processing is to perform correction processing on the first identification image information; Performing second processing on the first processed image to obtain a second processed image, wherein the second processing is to perform image enhancement processing on the first processed image; And comparing the second processed image with image sample data in a picture library, setting a similarity threshold of the second processed image and the image sample data, and obtaining second object identification data. In another embodiment of the present application, the comparing the second processed image with the image sample data in the image library, setting a similarity threshold between the second processed image and the image sample data, and obtaining second article identification data includes: Respectively comparing the image characteristic information of the second processed image with the image characteristic information of each image sample data in the image library, selecting a first target image sample with similarity meeting a first setting condition from the image sample data in the image library, and selecting a second target image sample with similarity meeting a second setting condition from the image sample data in the image library; Determining an image structural feature di