A drug recycling classification method, device, computing device and storage medium based on OCR

Abstract

The present invention discloses a drug recycling classification method based on optical character recognition (OCR), comprising: obtaining an image of a drug box to be recycled, recording it as an image to be recognized, and initiating a recognition request; a server receiving the recognition request, extracting text lines in the image to be recognized through optical character recognition (OCR), and recording the position of each text line; the server obtaining text title information and corresponding text content information based on the positional relationship of the text lines; when the image to be recognized contains a product batch number, a production date, and an expiration date, recording the order of the text lines, matching the text title information with the text content information one by one, and obtaining the values ​​of the product batch number, production date, and expiration date; the server matching the text content information with the required text title information and obtaining a matching result; calculating the remaining validity period based on the value of the expiration date, and sending a corresponding control signal based on the remaining validity period. This method achieves accurate classification of drug recycling, facilitates subsequent processing of drugs, and improves the efficiency of drug recycling.

Description

Technical Field

[0001] The present invention relates to the field of drug recycling, and more specifically, to an OCR-based drug recycling classification method, device computing equipment and storage medium. Background Art

[0002] The careless disposal of discarded and expired medicines can easily cause environmental pollution and harm human health. Currently, most people already know that excess medicines should not be thrown away randomly but should be placed in the correct recycling points. Therefore, many cities have established medicine recycling points to help recycle excess medicines. Due to the special nature of medicines, different medicines require different professional destruction methods for medical supplies. Furthermore, it is wasteful to simply dispose of some unused medicines. However, current medicine recycling points do not organize and classify the recycled medicines, nor do they fully utilize the recycling value of the medicines.

[0003] Patent No. CN113139401A discloses a method, computing device, drug recycling device and medium for drug recycling. The drug recycling device obtains a first image including a code indicating a drug identification on a medicine box to be recycled, identifies the code from the first image to obtain the drug identification, obtains a second image of the medicine box from a server based on the drug identification, generates a first hash value of the second image, obtains multiple third images including the medicine box to be recycled, generates multiple second hash values ​​of the multiple third images; and determines the validity of the medicine box to be recycled based on the first hash value and the multiple second hash values ​​to avoid fraud.

[0004] The drugs recovered by the drug recovery device are not sorted, but are all piled together. For drugs that are still a long way from expiration, direct destruction results in a huge waste of resources. Expired drugs need to be destroyed. Due to the different properties of different drugs, drugs with different medicinal properties require different treatment methods. The drugs recovered by the drug recovery device need to be manually sorted, which increases the subsequent drug recovery and sorting work. The relevant personnel can only manually sort when performing the recovery work, which is a large workload and increases the cost of drug recovery. Therefore, the drugs are sorted and recycled, and the drugs to be inspected are separated from the drugs to be processed. The drugs to be inspected are recycled, and after the inspection is completed, they can be promoted for processing, which greatly reduces the waste of drugs. The sorting of processed drugs reduces the workload of the recycling personnel and improves their work efficiency. Summary of the Invention

[0005] In view of the shortcomings of the above-mentioned technologies, the purpose of the present invention is to provide an OCR-based drug recycling classification method, device, computing equipment and storage medium to classify recycled drugs.

[0006] The present invention provides an OCR-based drug recycling classification method, which effectively solves the problems of serious waste of recycled drugs, difficulty in classifying recycled drugs by recycling personnel, and low work efficiency.

[0007] A drug recycling classification method based on OCR, comprising:

[0008] Obtain an image of the medicine box to be recycled, record it as the image to be identified, and initiate an identification request;

[0009] The server receives the recognition request, extracts text lines from the image to be recognized through OCR recognition, and records the position of each text line, wherein the text line includes text title information and text content information; the server obtains the text title information and its corresponding text content information based on the positional relationship of the text lines;

[0010] When the image to be identified contains a product batch number, production date, and expiration date, the text line order is recorded, and the text title information and text content information are matched one by one to obtain the values ​​of the product batch number, production date, and expiration date;

[0011] The server matches the text content information with the required text title information and obtains a matching result, which includes the drug name, properties, and packaging for determining the drug category.

[0012] The remaining validity period is calculated based on the value of the expiration date, and it is determined whether the remaining validity period falls within a first threshold. If so, a first control signal is issued to judge the recovered drug as a drug to be inspected and placed in a first drug area; if not, a second control signal is issued to judge the recovered drug as a drug to be processed and placed in a second drug area; the first control signal is used to control the recovery drug to be placed in the first drug area; the second control signal is used to control the processed drug to be placed in the second drug area;

[0013] Preferably, the image to be identified includes six views of the packaging of the medicine to be recycled;

[0014] Recording the position of each text line also includes:

[0015] Establish a rectangular coordinate system and record the coordinates of the text title information as (x i ,y i ), i = 1, 2, 3...; the coordinates of the text content information are recorded as (a j , b j ), j=1, 2, 3, ..., i and j refer to the number of text lines, and the coordinate point refers to the midpoint of the text line.

[0016] Preferably, when y j Equal to b j When |ai -x i | is equal to the second threshold, the text content information is after the text title information, the text title information and its corresponding text content information are obtained; when y j Not equal to b j When the image to be recognized has the product batch number, production date and expiration date, sort the text lines and put x i Same and x i i The first text title information, the second text title information, and the third text title information are recorded as x i Same and x i >a i The first text content information, the second text content information and the third text content information are recorded as follows: the first text title information corresponds to the first text content information, the second text title information corresponds to the second text content information, and the third text title information corresponds to the third text content information.

[0017] Preferably, the server matches the text content information with the required text title information, further comprising:

[0018] If the text content information matches the text title information, the remaining validity period of the drug to be recalled is calculated based on the matching result;

[0019] If the text content information does not match the text title information, the match is re-identified.

[0020] Preferably, it also includes: the matching result includes the drug's expiration date, expiration date, production date, drug name, properties and packaging; the remaining validity period is calculated based on the expiration date and the current date.

[0021] Preferably, it also includes:

[0022] The processed drugs continue to be classified, and the category to which the processed drugs belong is determined based on the drug name, properties and packaging; the category to which the processed drugs belong includes at least one category; the second control signal includes at least one sub-code, and the sub-code is used to determine the classification category to which the drugs to be processed belong, and the sub-code corresponds one-to-one to the classification category of the drugs to be processed.

[0023] A drug recovery and classification device based on OCR, comprising:

[0024] An image recognition unit acquires an image of the medicine box to be recycled, records it as an image to be recognized, and initiates a recognition request; the image to be recognized includes six views of the medicine box to be recycled;

[0025] ​An information extraction unit, in which a server receives the recognition request, extracts text lines from the image to be recognized through OCR recognition, and records the position of each text line, wherein the text line includes text title information and text content information; the server obtains the text title information and its corresponding text content information based on the positional relationship of the text lines; when the image to be recognized contains a product batch number, production date, and expiration date, the server records the order of the text lines, and associates the text title information with the text content information one-to-one to obtain the values ​​of the product batch number, production date, and expiration date;

[0026] The matching unit matches the text content information with the required text title information and obtains a matching result, which includes the drug name, properties, and packaging for determining the drug category;

[0027] The classification unit calculates the remaining validity period based on the value of the expiration date and determines whether the remaining validity period belongs to the first threshold value. If so, a first control signal is issued to judge the recovered drug as a drug to be inspected and put into the first drug area; if not, a second control signal is issued to judge the recovered drug as a processed drug and put into the second drug area.

[0028] Preferably, the matching unit further includes:

[0029] If the text content information matches the text title information, the remaining validity period of the drug to be recalled is calculated based on the matching result;

[0030] If the text content information does not match the text title information, re-identify and match;

[0031] The classification unit also includes:

[0032] The category of the drug to be processed is determined according to the drug name, properties and packaging; the second control signal includes at least one subcode, which is used to determine the classification category of the drug to be processed, and the subcode corresponds one-to-one to the classification category of the drug to be processed.

[0033] A computing device comprising:

[0034] At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor so as to enable the at least one processor to perform the method according to any one of claims 1 to 6.

[0035] A non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause the computer to execute the method according to any one of claims 1 to 6.

[0036] The present invention has at least the following beneficial effects:

[0037] The drugs in the drug recovery device are automatically classified, and the drugs to be inspected and the processed drugs are classified in different ways, replacing the complicated classification work of the later recycling personnel. For the processed drugs, it not only greatly improves the correctness and accuracy of the drug classification, but also reduces the workload of the recycling personnel and improves the classification efficiency of the drugs to be recycled; for the drugs to be inspected, the drugs are verified and inspected after recycling, and then the drugs can be recycled to the pharmacy for promotion and processing, reducing the waste of drugs.

[0038] Other advantages, objectives and features of the present invention will be reflected in part through the following description, and in part will be understood by those skilled in the art through study and practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a basic flow chart of the OCR-based drug recycling classification method;

[0040] Figure 2 This is a detailed implementation flow chart of the drug delivery in Example 1;

[0041] Figure 3 The following schematically shows a structural block diagram of a device for drug recycling classification based on OCR according to an embodiment of the present disclosure;

[0042] Figure 4 The structure block diagram of the electronic device provided by the embodiment of the present disclosure is schematically shown; DETAILED DESCRIPTION

[0043] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all the embodiments. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0044] Example 1

[0045] This embodiment provides a drug recycling classification method based on OCR. Figure 1 As shown:

[0046] S1: Obtain an image of the medicine box to be recycled, record it as the image to be identified, and initiate an identification request;

[0047] Specifically, the image to be identified refers to the six views of the medicine box to be recycled, including the front view, rear view, left view, right view, top view, and top view of the medicine box to be recycled, which facilitates the extraction of more information about the medicine to be recycled, improves the recognition accuracy, and reduces the misjudgment rate.

[0048] S2: The server receives the recognition request, extracts text lines from the image to be recognized through OCR recognition, and records the position of each text line, wherein the text line includes text title information and text content information; the server obtains the text title information and its corresponding text content information based on the positional relationship of the text lines;

[0049] Specifically, the text title information and the location of the text title information in the image to be extracted are determined by first extracting all the text title information and the coordinate location of each text title information in the image to be identified through optical character recognition (OCR), and then extracting the text content information in the image to be identified through OCR. The text content information refers to the specific and detailed drug information on a medicine box, for example, "Packaging: Composite film packaging, 5g per bag, 9 bags per box" on the medicine box, where "Packaging" is the text title information and "Composite film packaging, 5g per bag, 9 bags per box" is the text content information.

[0050] In actual applications, the text title information is generally fixed. For example, the text title information on a medicine box generally includes the drug name, ingredients, properties, functions, specifications, usage and dosage, expiration date, production date, product batch number, approval number, etc.

[0051] Therefore, after first extracting all text title information and the coordinate position of each text title information from the image to be identified through OCR recognition, the text content information of the image to be identified corresponding to each text title information can be quickly and conveniently extracted using OCR recognition technology based on the coordinate position of each text title information. For example, if the text content information is "6 tablets / plate x 5 plates / box", the text title information corresponding to this text content information is "Specifications"; if the text content information is "Bi Ling Wei Tong Granules", the text title information corresponding to this text content information is "Drug Name".

[0052] S3: When the image to be recognized contains a product batch number, production date, and expiration date, record the text line order, match the text title information with the text content information one by one, and obtain the product batch number, production date, and expiration date values;

[0053] Specifically, recording the position of each text line also includes: establishing a rectangular coordinate system, recording the coordinates of the text title information as (x i ,y i ), i = 1, 2, 3...; the coordinates of the text content information are recorded as (aj , b j ), j=1, 2, 3, ..., i and j refer to the number of text lines, and the coordinate point refers to the midpoint of the text line.

[0054] When y j Equal to b j When |a i -x i | is equal to the second threshold, the text content information is after the text title information, and the text title information and its corresponding text content information are obtained; the second threshold refers to the distance between the text title information and the text content information;

[0055] When y j Not equal to b j When the image to be recognized has the product batch number, production date and expiration date, sort the text lines and put x i Same and x i i The first text title information, the second text title information, and the third text title information are recorded as x i Same and x i >a i The first text content information, the second text content information, and the third text content information are recorded as follows: the first text title information corresponds to the first text content information, the second text title information corresponds to the second text content information, and the third text title information corresponds to the third text content information. For example, if the first text title information is the product batch number, the second text title information is the production date, and the third text title information is the expiration date, then the first text content information is the product batch number value, the second text content information is the production date value, and the third text content information is the expiration date value. Therefore, the expiration date value is obtained.

[0056] S4: The server matches the text content information with the required text title information and obtains a matching result, which includes the drug name, properties, and packaging for determining the drug category;

[0057] Specifically, the required text title information refers to the text information title that is desired to be obtained in advance, and the remaining validity period and drug category are obtained through the text title information, such as the production date, expiration date, expiration date, drug name, packaging, properties and other text title information.

[0058] Taking the Biling Weitong Granules medicine box image as an example, we extract some text title information and text content information, including:

[0059] ​Name of the drug: Bi Ling Wei Tong Granules; Ingredients: Piper serrata, Chuan Rou Zi, Vinegar Corydalis, Wine Rhubarb, Coptis chinensis, Evodia rutaecarpa, Vinegar Cyperus rotundus, Citron, Citron, Octopus octopus, Calcined Wa Bang Zi; Properties: This product is brown to dark brown granules with a bitter taste; Functions and indications: Promotes qi and blood circulation, harmonizes the stomach and relieves pain, used for epigastric pain caused by qi stagnation and blood stasis, and chronic gastritis with the above symptoms; Specifications: Each bag is equivalent to 20 grams of the decoction; Usage and dosage: Take with boiled water, 5 grams at a time, 3 times a day; Packaging: Composite film packaging, 5g per bag, 9 bags per box; Validity period: 18 months; Approval number: National Medicine Standard Z19990069; Production date: 2022 / 09 / 18; Valid until: 2024 / 02.

[0060] The required text title information includes: drug name, production date, expiration date, properties, and packaging.

[0061] The extracted text title information and text content information are matched with the required text title information. If the match is successful and the required text title information corresponds to the text content information one by one, a matching result is obtained, including:

[0062] Drug Name: Piling Weitong Granules; Production Date: September 18, 2022; Expiration Date: February 2024; Validity Period: 18 months; Properties: Brown to pale brown granules with a bitter taste; Packaging: Composite film packaging, 5g per bag, 9 bags per box;

[0063] Calculate the remaining validity period of the drug to be recalled based on the matching result, where the remaining validity period = validity period - current date;

[0064] In another case, if the required text title information includes: drug name, adverse reactions, and company name, the extracted text title information and text content information are matched with the required text title information. If the match fails, "adverse reactions" and "company name" do not match the corresponding text title information and text content information, repeat S1, S2, and S3 until a match is successful.

[0065] S5: Calculate the remaining validity period based on the value of the expiration date, and determine whether the remaining validity period falls within a first threshold. If so, issue a first control signal to determine the recovered drug as a drug to be inspected and placed in a first drug area; if not, issue a second control signal to determine the recovered drug as a drug to be processed and placed in a second drug area; the first control signal is used to control the recovery drug to be placed in the first drug area; the second control signal is used to control the processed drug to be placed in the second drug area;

[0066] Specifically, such as Figure 2 As shown:

[0067] S501: After calculating the remaining validity period, determine whether the remaining validity period falls within a first threshold. If so, execute S502; otherwise, execute S503;

[0068] Specifically, the first threshold may be 90 or above.

[0069] S502: Sending a first control signal to put the medicine into the first medicine area;

[0070] S503: Determine the category of the drug based on the key text in the matching result;

[0071] Specifically, the drug category to be processed can be determined based on the matched drug name, properties, and packaging. These include tablets, pills, capsules, topical liquids, oral solutions, and other liquid preparations, ointments, sprays, injections, and other treatment categories. Different destruction methods are used for different drug categories.

[0072] S504: Sending a second control signal to put the drugs to be processed into the corresponding second drug classification area.

[0073] Specifically, the second control signal includes at least one sub-code, which is used to determine the classification category of the drugs to be processed, and the sub-code corresponds one-to-one to the classification category of the drugs to be processed; for example, tablets, pills, capsules and other dosage forms of drugs correspond to sub-code A, external medicines such as oral liquids correspond to sub-code B, ointment preparations correspond to sub-code C, sprays correspond to sub-code D, injections and aqueous solutions correspond to sub-code E, and other categories of drugs to be processed correspond to sub-code F; since different drugs have different properties, drugs with different medicinal properties require different processing methods, so classifying the drugs to be processed facilitates the subsequent processing of the drugs to be processed and reduces the workload of recycling personnel.

[0074] The OCR-based drug recycling classification method provided in this embodiment scans and identifies drugs to be recycled, and classifies drugs to be inspected and drugs to be processed differently, which greatly improves the utilization rate of drugs and the accuracy of drug recycling classification, facilitates the subsequent processing of drugs, reduces the workload of recycling personnel, and improves the efficiency of drug recycling.

[0075] Example 2

[0076] An OCR-based drug recovery classification device 200, characterized by comprising:

[0077] The image recognition unit 201 obtains an image of the medicine box to be recycled, records it as an image to be recognized, and initiates a recognition request; the image to be recognized includes six views of the medicine box to be recycled;

[0078] Information extraction unit 202: The server receives the recognition request, extracts text lines from the image to be recognized through OCR recognition, and records the position of each text line, wherein the text line includes text title information and text content information. The server obtains the text title information and its corresponding text content information based on the positional relationship of the text lines. When the image to be recognized contains a product batch number, production date, and expiration date, the server records the order of the text lines, associates the text title information with the text content information one-to-one, and obtains the values ​​of the product batch number, production date, and expiration date.

[0079] Matching unit 203, the server matches the text content information with the required text title information and obtains a matching result, which includes the drug name, properties, and packaging for determining the drug category;

[0080] Specifically, if the text content information matches the text title information, the remaining validity period of the drug to be recalled is calculated based on the matching result;

[0081] If the text content information does not match the text title information, re-identify and match;

[0082] The classification unit 204 calculates the remaining validity period of the drugs to be recovered based on the matching results; determines whether the remaining validity period falls within a first threshold value. If so, a first control signal is issued to judge the drugs to be recovered as drugs to be inspected and put them into the first drug area; if not, a second control signal is issued to judge the drugs to be recovered as drugs to be processed and put them into the second drug area.

[0083] Specifically, the category of the drug to be processed is determined based on the drug name, properties, and packaging; the second control signal includes at least one subcode, the subcode is used to determine the classification category of the drug to be processed, and the subcode corresponds to the classification category of the drug to be processed;

[0084] Example 3

[0085] Computer-readable storage medium can be a tangible device that can keep and store the instruction used by the instruction execution device, for example, can be but not limited to electrical storage device, magnetic storage device, optical storage device, electromagnetic storage device, semiconductor storage device or above-mentioned any suitable combination.The more specific example of computer-readable storage medium comprises: portable computer disk, hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), digital versatile disk (DVD), mechanical encoding device, for example, punch card or groove inner convex structure and above-mentioned any suitable combination that are stored thereon with instruction.Computer-readable storage medium used here is not interpreted as instantaneous signal itself, such as radio wave or other freely propagating electromagnetic wave, electromagnetic wave propagated by waveguide or other transmission medium or electric signal transmitted by wire.

[0086] A computing device 100. The computing device includes, but is not limited to, a memory 102 and a processor 101. The processor 101 is connected to the memory 102 via a bus 104, and a database 106 is used to store data for facilitating access by a server 105. The computing device 100 also includes an access device 103, which enables the computing device to communicate via one or more networks. Examples of these networks include a public switched telephone network (PSTN), a local area network (LAN), a wide area network (WAN), or a combination of communication networks such as the Internet. The access device may include one or more of any type of network interface (e.g., a network interface card (NIC)) of wired or wireless type, such as a wireless local area network (WLAN) wireless interface, an Ethernet interface, a universal serial bus (USB) interface, a cellular network interface, a near field communication (NFC) interface, and the like.

[0087] The computing device 100 may be any type of stationary or mobile computing device, including a mobile computer or mobile computing device (e.g., a tablet computer, a laptop computer, a netbook computer, etc.), a mobile phone (e.g., a smartphone), or a stationary computing device such as a desktop computer or PC. The computing device 100 may also be a mobile or stationary server.

[0088] While various embodiments of the present disclosure have been described above, the foregoing description is intended to be illustrative, non-exhaustive, and not limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is selected to best explain the principles of the embodiments, their practical applications, or technical improvements to existing technologies, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A drug recycling classification method based on OCR, characterized in that: include: Obtain an image of the medicine box to be recycled, record it as the image to be identified, and initiate an identification request; The server receives the recognition request, extracts text lines from the image to be recognized through OCR recognition, and records the position of each text line, wherein the text line includes text title information and text content information; the server obtains the text title information and its corresponding text content information based on the positional relationship of the text lines; When the image to be identified contains a product batch number, production date, and expiration date, the text line order is recorded, and the text title information and text content information are matched one by one to obtain the values ​​of the product batch number, production date, and expiration date; The server matches the text content information with the required text title information and obtains a matching result, which includes the drug name, properties, and packaging for determining the drug category. The remaining validity period is calculated based on the value of the expiration date, and it is determined whether the remaining validity period belongs to the first threshold value. If so, a first control signal is issued to judge the recovered drug as a drug to be inspected and put into the first drug area; if not, a second control signal is issued to judge the recovered drug as a drug to be processed and put into the second drug area; the first control signal is used to control the recovery drug to be put into the first drug area; the second control signal is used to control the processed drug to be put into the second drug area.

2. The OCR-based drug recovery classification method according to claim 1, characterized in that: The image to be identified includes six views of the packaging of the medicine to be recycled; Recording the position of each text line also includes: Establish a rectangular coordinate system and record the coordinates of the text title information as (x i ,y i ), i = 1, 2, 3...; the coordinates of the text content information are recorded as (a j , b j ), j=1, 2, 3, ..., i and j refer to the number of text lines, and the coordinate point refers to the midpoint of the text line.

3. The OCR-based drug recovery classification method according to claim 2, characterized in that: When y j Equal to b j When |a i -x i | is equal to the second threshold, the text content information is after the text title information, the text title information and its corresponding text content information are obtained; when y j Not equal to b j When the image to be recognized has the product batch number, production date and expiration date, sort the text lines and put x i Same and x i i The first text title information, the second text title information, and the third text title information are recorded as x i Same and x i >a i The first text content information, the second text content information and the third text content information are recorded as follows: the first text title information corresponds to the first text content information, the second text title information corresponds to the second text content information, and the third text title information corresponds to the third text content information.​ 4. The OCR-based drug recovery classification method according to claim 1, characterized in that: The server matches the text content information with the required text title information, and also includes: If the text content information matches the text title information, the remaining validity period of the drug to be recalled is calculated based on the matching result; If the text content information does not match the text title information, the match is re-identified.

5. The OCR-based drug recovery classification method according to claim 1, characterized in that: Also includes: The matching result includes the drug's expiration date, expiration date, production date, drug name, properties and packaging; the remaining validity period is calculated based on the expiration date and the current date.

6. The OCR-based drug recovery classification method according to claim 1, characterized in that: Also includes: Continue to classify the drugs to be processed and determine the category to which the drugs to be processed belong based on the drug name, properties and packaging; The category to which the drugs to be processed belong includes at least one category; the second control signal includes at least one subcode, the subcode is used to determine the classification category to which the drugs to be processed belong, and the subcode corresponds one-to-one to the classification category of the drugs to be processed.

7. A drug recovery and classification device based on OCR, characterized in that: include: An image recognition unit acquires an image of the medicine box to be recycled, records it as an image to be recognized, and initiates a recognition request; the image to be recognized includes six views of the medicine box to be recycled; An information extraction unit, in which a server receives the recognition request, extracts text lines from the image to be recognized through OCR recognition, and records the position of each text line, wherein the text line includes text title information and text content information; the server obtains the text title information and its corresponding text content information based on the positional relationship of the text lines; when the image to be recognized contains a product batch number, production date, and expiration date, the server records the order of the text lines, and associates the text title information with the text content information one-to-one to obtain the values ​​of the product batch number, production date, and expiration date; The matching unit matches the text content information with the required text title information and obtains a matching result, which includes the drug name, properties, and packaging for determining the drug category; The classification unit calculates the remaining validity period based on the value of the expiration date and determines whether the remaining validity period belongs to the first threshold value. If so, a first control signal is issued to judge the recovered drug as a drug to be inspected and put into the first drug area; if not, a second control signal is issued to judge the recovered drug as a drug to be processed and put into the second drug area.

8. The OCR-based drug recovery and classification device according to claim 7, characterized in that: The matching unit further includes: If the text content information matches the text title information, the remaining validity period of the drug to be recalled is calculated based on the matching result; If the text content information does not match the text title information, re-identify and match; The classification unit also includes: The category of the drug to be processed is determined according to the drug name, properties and packaging; the second control signal includes at least one subcode, which is used to determine the classification category of the drug to be processed, and the subcode corresponds one-to-one to the classification category of the drug to be processed.

9. A computing device, characterized in that include: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method according to any one of claims 1 to 6.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to cause the computer to execute the method according to any one of claims 1 to 6.

Images