Method and system for automatically identifying and counting medical instruments

Abstract

The invention provides a medical instrument automatic identification counting method and system, and the method comprises the steps: collecting the instrument image of each medical instrument through an image collection device when a plurality of to-be-counted medical instruments are placed on a weighing platform; identifying the instrument image of each medical instrument, and determining the instrument type of each medical instrument; based on the standard weight of the instrument type of each medical instrument and a weighing result obtained by weighing each medical instrument by the weighing platform, determining a counting result of each medical instrument; and outputting the counting result of each medical device. Through the automatic image recognition and weighing technology, the labor cost can be greatly reduced, the counting accuracy is improved, and omissions and errors in manual counting are avoided; a large number of medical instruments can be quickly processed, the counting efficiency is greatly improved, and the requirements in the aspects of equipment inventory, distribution and the like can be quickly responded.

Description

Technical Field

[0001] The present invention relates to the technical field of computer data processing, and in particular to a method and system for automatically identifying and counting medical devices. Background Art

[0002] With the rapid development of the medical industry, the types and quantity of medical devices are increasing, and the management and use of medical devices are becoming more complex. Accurate counting and classification of medical devices is crucial to the operation of hospitals, clinics and related supply chains, especially in the process of purchasing, storing, distributing and using devices. Ensuring the accuracy of the number of devices is the key to ensuring medical safety and improving work efficiency.

[0003] At present, the counting of medical devices generally relies on manual operation or traditional manual inventory. This method is not only inefficient, but also susceptible to human errors. In the process of managing a large number of batches of medical devices, manual counting errors are often difficult to avoid. In addition, manual counting also requires a lot of time and manpower costs, and cannot quickly respond to the needs of equipment inventory, distribution, etc., resulting in frequent out-of-stock, expired or inaccurate inventory problems in medical institutions during use. Summary of the invention

[0004] One of the purposes of the present invention is to provide a method for automatic identification and counting of medical devices. Through automated image recognition and weighing technology, it can greatly reduce labor costs, improve counting accuracy, and avoid omissions and errors in manual counting; it can quickly process a large number of medical devices, greatly improve counting efficiency, and can quickly respond to equipment inventory, distribution and other needs; image recognition technology can accurately identify various types of medical devices and match them with standard weights, so that the system can cope with various complex medical device packaging, further avoiding the tedious operations of manual classification and matching.

[0005] An embodiment of the present invention provides a method for automatically identifying and counting medical devices, comprising:

[0006] When a large number of medical devices to be counted are placed on the weighing platform, the image of each medical device is collected by the image acquisition device;

[0007] Identify the device image of each medical device and determine the device type of each medical device;

[0008] Determine the counting result of each medical device based on the standard weight of the device type of each medical device and the weighing result of each medical device weighed by the weighing platform;

[0009] Output the counting results of each medical device.

[0010] Optionally, the identifying the device image of each medical device and determining the device type of each medical device includes:

[0011] Matching the device image of each medical device with a plurality of standard device images in a standard device image library;

[0012] When the matching is successful, the standard instrument types of the multiple standard instrument images that match the matching are used as the instrument type of each medical instrument.

[0013] Optionally, the counting result of each medical device is determined based on the standard weight of the device type of each medical device and the weighing result of each medical device weighed by the weighing platform, including:

[0014] Randomly combining the standard weights of device types of each medical device to obtain multiple weight sets;

[0015] Calculate the sum of the weights of each weight set;

[0016] The counting result of each medical device is determined based on the weight and the number of standard weights in the weight set with the smallest difference from the weighing result and the corresponding device type.

[0017] Optionally, the collecting of the device images of each medical device by an image collection device includes:

[0018] Obtaining the placement trajectory of each medical device on the weighing platform;

[0019] Based on the placement order of each medical device on the weighing platform, the placement trajectories corresponding to each medical device are sorted to obtain a placement trajectory sequence;

[0020] Based on the placement trajectory sequence, plan the shooting trajectory sequence;

[0021] Based on the shooting trajectory sequence, the image acquisition device is controlled to acquire the instrument image of each medical instrument.

[0022] Optionally, planning a shooting trajectory sequence based on the placement trajectory sequence includes:

[0023] Based on the local sequence partitioning constraint, multiple local sequences are divided in the placement trajectory sequence;

[0024] Traverse each local sequence in turn according to the sequence order of each local sequence in the placement trajectory sequence;

[0025] During each traversal, a local shooting trajectory sequence of the traversed local sequence is planned based on the shooting trajectory planning constraints;

[0026] After traversing each local sequence, the local shooting trajectory sequence planned by each traversal is spliced ​​in the order of traversal to obtain a shooting trajectory sequence;

[0027] Among them, the local sequence partition constraints include:

[0028] Constraint 1: The same local sequence contains at least N consecutive placement trajectories in the placement trajectory sequence; N is a positive integer;

[0029] And,Constraint 2, the placement trajectories in different local sequences are different;

[0030] And, constraint three, in the same local sequence, there is a shooting flip space between the first track segment of the first proportion behind the former and the second track segment of the second proportion in front of the latter in the adjacent placement tracks; wherein the shooting flip space includes: a spherical space between the first plane where the front, middle and last three points on the first track segment are located and the second plane where the front, middle and last three points on the second track segment are located, with a radius length of the radius value, and the vertical distances from the center of the sphere to the first plane and the second plane respectively not exceeding the distance threshold;

[0031] Among them, the shooting trajectory planning constraints include:

[0032] Constraint 4: When the image acquisition device shoots along the i-th shooting trajectory in the shooting trajectory sequence, it can continuously change all standard shooting postures to shoot the medical device placed on the weighing platform along the i-th placement trajectory in the traversed local sequence; i = 1, 2, ..., N; N is the total number of shooting trajectories in the shooting trajectory sequence or the total number of placement trajectories in the traversed local sequence; i takes the same value each time;

[0033] And, constraint 5: the end point of the i-th shooting trajectory in the shooting trajectory sequence is the center of the shooting flip space of the i-th group of adjacent placement trajectories in the traversed local sequence; i takes the same value each time;

[0034] And, constraint six, when the image acquisition device shoots along the i-th and i+1-th shooting trajectories in the shooting trajectory sequence, at least one lens flip is performed in the shooting flip space of the i-th group of adjacent placement trajectories in the traversed local sequence; i takes the same value each time.

[0035] An embodiment of the present invention provides a medical device automatic identification and counting system, comprising:

[0036] A collection module, used for collecting images of each medical device through an image collection device when a large number of medical devices to be counted are placed on the weighing platform;

[0037] An identification module, used to identify the device image of each medical device and determine the device type of each medical device;

[0038] A determination module, for determining a counting result of each medical device based on a standard weight of a device type of each medical device and a weighing result of each medical device weighed by a weighing platform;

[0039] The output module is used to output the counting results of each medical device.

[0040] Optional,

[0041] The recognition module recognizes the device image of each medical device and determines the device type of each medical device, including:

[0042] Matching the device image of each medical device with a plurality of standard device images in a standard device image library;

[0043] When the matching is successful, the standard instrument types of the multiple standard instrument images that match the matching are used as the instrument type of each medical instrument.

[0044] Optionally, the determination module determines the counting result of each medical device based on the standard weight of the device type of each medical device and the weighing result of each medical device weighed by the weighing platform, including:

[0045] Randomly combining the standard weights of device types of each medical device to obtain multiple weight sets;

[0046] Calculate the sum of the weights of each weight set;

[0047] The counting result of each medical device is determined based on the weight and the number of standard weights in the weight set with the smallest difference from the weighing result and the corresponding device type.

[0048] Optionally, the acquisition module acquires the device image of each medical device through an image acquisition device, including:

[0049] If each medical device is a package, obtain the placement track of each medical device package placed on the weighing platform;

[0050] Based on the placement order of each medical device on the weighing platform, the placement trajectories corresponding to each medical device are sorted to obtain a placement trajectory sequence;

[0051] Based on the placement trajectory sequence, plan the shooting trajectory sequence;

[0052] Based on the shooting trajectory sequence, the image acquisition device is controlled to acquire the instrument image of each medical instrument.

[0053] Optionally, the acquisition module plans a shooting trajectory sequence based on the placement trajectory sequence, including:

[0054] Based on the local sequence partitioning constraint, multiple local sequences are divided in the placement trajectory sequence;

[0055] Traverse each local sequence in turn according to the sequence order of each local sequence in the placement trajectory sequence;

[0056] During each traversal, a local shooting trajectory sequence of the traversed local sequence is planned based on the shooting trajectory planning constraints;

[0057] After traversing each local sequence, the local shooting trajectory sequence planned by each traversal is spliced ​​in the order of traversal to obtain a shooting trajectory sequence;

[0058] Among them, the local sequence partition constraints include:

[0059] Constraint 1: The same local sequence contains at least N consecutive placement trajectories in the placement trajectory sequence; N is a positive integer;

[0060] And,Constraint 2, the placement trajectories in different local sequences are different;

[0061] And, constraint three, in the same local sequence, there is a shooting flip space between the first track segment of the first proportion behind the former and the second track segment of the second proportion in front of the latter in the adjacent placement tracks; wherein the shooting flip space includes: a spherical space between the first plane where the front, middle and last three points on the first track segment are located and the second plane where the front, middle and last three points on the second track segment are located, with a radius length of the radius value, and the vertical distances from the center of the sphere to the first plane and the second plane respectively not exceeding the distance threshold;

[0062] Among them, the shooting trajectory planning constraints include:

[0063] Constraint 4: When the image acquisition device shoots along the i-th shooting trajectory in the shooting trajectory sequence, it can continuously change all standard shooting postures to shoot the medical device package placed on the weighing platform along the i-th placement trajectory in the traversed local sequence; i = 1, 2, ..., N; N is the total number of shooting trajectories in the shooting trajectory sequence or the total number of placement trajectories in the traversed local sequence; i takes the same value each time;

[0064] And, constraint 5: the end point of the i-th shooting trajectory in the shooting trajectory sequence is the center of the shooting flip space of the i-th group of adjacent placement trajectories in the traversed local sequence; i takes the same value each time;

[0065] And, constraint six, when the image acquisition device shoots along the i-th and i+1-th shooting trajectories in the shooting trajectory sequence, at least one lens flip is performed in the shooting flip space of the i-th group of adjacent placement trajectories in the traversed local sequence; i takes the same value each time.

[0066] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0067] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0069] Figure 1 A schematic diagram of a method for automatically identifying and counting medical devices in an embodiment of the present invention;

[0070] Figure 2 A schematic diagram of a specific application flow of the method for automatically identifying and counting medical devices in an embodiment of the present invention;

[0071] Figure 3 Schematic diagram of a medical device automatic identification and counting system in an embodiment of the present invention. DETAILED DESCRIPTION

[0072] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0073] The embodiment of the present invention provides a method for automatically identifying and counting medical devices, such as Figure 1 As shown, including:

[0074] S1. When a large number of medical devices to be counted are placed on the weighing platform, an image acquisition device is used to acquire images of each medical device;

[0075] S2. Identify the device image of each medical device and determine the device type of each medical device;

[0076] S3, determining the counting result of each medical device based on the standard weight of the device type of each medical device and the weighing result of each medical device weighed by the weighing platform;

[0077] S4. Output the counting results of each medical device.

[0078] like Figure 2As shown, first, when a large number of medical devices to be counted (such as a pile of dental handpieces, a periodontal probe) are placed on a weighing platform, the image acquisition device (such as a high-resolution camera or a visual sensor) is used to collect the device image of each medical device. During the collection, the image acquisition device will shoot the device image of each medical device from the top of the weighing platform downward. However, when there are a large number of certain medical devices, it is impossible to count based on the device image (such as a pile of dental handpieces, a periodontal probe). Therefore, for such cases, the medical devices cannot be counted based on the device image. The device image contains the appearance information of the medical device; the system uses image recognition technology to process the collected image. Based on visual recognition technology, the system can analyze the features of the device image and identify the device type of each medical device, which may include the brand, model, specification, etc. of the device for further counting analysis. After identifying the type of medical device, the system calculates according to the standard weight of each device type. The standard weight information can be stored in a database and corresponds to the device image of each medical device. Once the system knows the type and standard weight of the device, it can use the weighing data provided by the weighing platform to count. The total weight of each device will be accurately measured by the weighing platform. The system will calculate the number of devices in the package by comparing the ratio of this weight to the standard weight. For example, if the total weight of a medical device matches the standard weight, it means that the device contains one unit. The system will calculate according to the standard weight ratio to get an accurate count result. The system will output the count results of each medical device. These results will be displayed on a visual interface for staff to check, record or follow-up operations.

[0079] This application uses automated image recognition and weighing technology to greatly reduce labor costs, improve counting accuracy, and avoid omissions and errors in manual counting; it can quickly process a large number of medical devices, greatly improve counting efficiency, and quickly respond to equipment inventory, distribution and other needs; image recognition technology can accurately identify various types of medical devices and match them with standard weights, allowing the system to cope with various complex medical device packaging, further avoiding the tedious operations of manual classification and matching.

[0080] In one embodiment, the step S2, identifying the device image of each medical device and determining the device type of each medical device, includes:

[0081] S21, matching the device image of each medical device with a plurality of standard device images in a standard device image library;

[0082] S22: When the matching is successful, the standard device type of the multiple standard device images that match the matching is used as the device type of each medical device.

[0083] The embodiment of the present invention matches the device image of each medical device with a plurality of standard device images in a standard device image library, and uses image matching technology to compare with the standard device images, thereby determining the specific device type of each medical device.

[0084] In one embodiment, the step S3, determining the counting result of each medical device based on the standard weight of the device type of each medical device and the weighing result of each medical device weighed by the weighing platform, includes:

[0085] S31, randomly combining the standard weights of the device types of various medical devices to obtain multiple weight sets;

[0086] S32, calculating the weight sum of each weight set;

[0087] S33. Determine the counting result of each medical device based on the weight and the number of standard weights in the weight set with the smallest difference from the weighing result and the corresponding device type.

[0088] By randomly combining the standard weights of each medical device, multiple possible weight sets are obtained, and the total weight of these combinations is calculated; then, the calculated total weight is compared with the actual weighing result, and the combination with the smallest difference is selected to determine the specific number and type of each medical device.

[0089] For example, suppose a large number of medical devices weigh 500 grams and contain different types of medical devices. The system knows the standard weights of several devices as follows:

[0090] Equipment A: 100g;

[0091] Equipment B: 150g;

[0092] Equipment C: 200 g;

[0093] The system will randomly combine these standard weights, for example:

[0094] Combination 1: 2 apparatus A and 1 apparatus B, total weight is 100g × 2 + 150g × 1 = 350g;

[0095] Combination 2: 1 device A and 2 devices C, with a total weight of 100 g × 1 + 200 g × 2 = 500 g;

[0096] Combination 3: 1 apparatus B and 1 apparatus C, with a total weight of 150 g × 1 + 200 g × 1 = 350 g;

[0097] The system then calculates the total weight of these combinations and finds that the sum of the weights of combination 2 (500 grams) is exactly the same as the actual weighing result. Therefore, the system confirms that the medical device contains 1 device A and 2 devices C, and finally determines the counting result of the medical device as: 1 device A and 2 devices C.

[0098] In one embodiment, in S1, collecting the device image of each medical device by an image acquisition device includes:

[0099] S11. If each medical device is a package, obtain a placement track of each medical device package placed on the weighing platform;

[0100] In S11, in order to further automate and unmanned the automatic identification technology of medical devices, a robot can be set to place the medical device packages on the weighing platform. After the setting, the robot will plan in real time how to place each medical device package on the weighing platform, that is, generate a planned placement trajectory for each medical device package in the future;

[0101] S12, based on the placement order of each medical device package on the weighing platform, sorting the placement tracks of each medical device package to obtain a placement track sequence;

[0102] In S12, when planning, the robot will also plan the order in which each medical device package is placed on the weighing platform; the placement trajectory of each medical device package is sorted according to the placement order to obtain a placement trajectory sequence;

[0103] S13, planning a shooting trajectory sequence based on the placement trajectory sequence;

[0104] In S13, the placement trajectory sequence represents the movement of each medical device package placed on the weighing platform. The image acquisition device needs to acquire the packaging image of each medical device package. Therefore, the shooting trajectory sequence can be planned based on the placement trajectory sequence. The shooting trajectory sequence indicates how the image acquisition device moves to complete the acquisition of the packaging image of each medical device package.

[0105] S14, based on the shooting trajectory sequence, controlling the image acquisition device to acquire the packaging image of each medical device package;

[0106] In S14, the image acquisition device is finally controlled based on the shooting trajectory sequence, that is, the acquisition of the packaging image is completed;

[0107] S15. Using the type label image on the packaging image as the device image of each medical device.

[0108] When the embodiment of the present invention is applied, the staff uses non-fixed packaging to pack a large number of different medical devices. When packing, there can be multiple medical devices in the same medical device package. After the packaging is completed, the exclusive label of the packaged medical device is affixed to the outer packaging of the package. Since the packaging of medical devices is not fixed, the staff will only affix the device type label on one side when packing. During the transportation to the identification technology site and placement on the weighing platform, the orientation of the label surface will change many times. It is very likely that after being placed on the weighing platform, the label surface cannot be imaged because it is facing downward, and it may also be blocked by other things and cannot be imaged. The embodiment of the present invention can solve this problem. In the process of placing each medical device on the weighing platform, the image acquisition device is controlled to move and collect the device image of each medical device through the shooting trajectory sequence planned based on the placement trajectory sequence, so that the image of the device type label on each medical device can be collected, so that the device type can be determined, which greatly improves the applicability of the system.

[0109] When multiple medical device packages to be counted are placed on the weighing platform, the package image is collected to determine the label and the type of device; after identifying the type of medical device, the system calculates according to the standard weight of each device type. The standard weight information can be stored in the database and corresponds to the package image of each medical device. Once the system knows the type and standard weight of the device, it can use the weighing data provided by the weighing platform to count. The total weight of each package will be accurately measured by the weighing platform, and the system will calculate the number of devices in the package by comparing the ratio of this weight to the standard weight; for example, if the total weight of a package is consistent with the standard weight, it means that the package contains one unit of device. If the weight deviation is large, it may mean that the package contains multiple devices. The system will calculate according to the standard weight ratio to obtain an accurate counting result. The system will output the counting result of each medical device package. By randomly combining the standard weights of each medical device package, multiple possible weight sets are obtained, and the total weight of these combinations is calculated; then, the calculated total weight is compared with the actual weighing result, and the combination with the smallest difference is selected to determine the specific number and type of devices in each package.

[0110] For example, suppose a medical device package weighs 500 grams and contains different types of medical devices. The system knows the standard weights of several devices as follows:

[0111] Equipment A: 100g;

[0112] Equipment B: 150g;

[0113] Equipment C: 200 g;

[0114] The system will randomly combine these standard weights, for example:

[0115] Combination 1: 2 apparatus A and 1 apparatus B, total weight is 100g × 2 + 150g × 1 = 350g;

[0116] Combination 2: 1 device A and 2 devices C, with a total weight of 100 g × 1 + 200 g × 2 = 500 g;

[0117] Combination 3: 1 apparatus B and 1 apparatus C, with a total weight of 150 g × 1 + 200 g × 1 = 350 g;

[0118] The system then calculates the total weight of these combinations and finds that the sum of the weight of combination 2 (500 grams) is exactly the same as the actual weighing result. Therefore, the system confirms that the package contains 1 device A and 2 devices C, and finally determines the count result of the package as: 1 device A and 2 devices C.

[0119] In one embodiment, the step S13, planning a shooting trajectory sequence based on the placement trajectory sequence, includes:

[0120] S131, dividing a plurality of local sequences in the placement trajectory sequence based on the local sequence division constraint;

[0121] S132, traversing each local sequence in sequence according to the sequence order of each local sequence in the placement trajectory sequence;

[0122] S133, planning a local shooting trajectory sequence of the traversed local sequence based on the shooting trajectory planning constraint during each traversal;

[0123] S134, after traversing each local sequence, the local shooting trajectory sequence planned by each traversal is spliced ​​in the order of traversal to obtain a shooting trajectory sequence;

[0124] In S134, when stitching, the shooting trajectory sequences obtained by traversing the local sequence earlier are arranged closer to the front for stitching;

[0125] Among them, the local sequence partition constraints include:

[0126] Constraint 1: The same local sequence contains at least N consecutive placement trajectories in the placement trajectory sequence; N is a positive integer;

[0127] And,Constraint 2, the placement trajectories in different local sequences are different;

[0128] And, constraint three, in the same local sequence, there is a shooting flip space between the first track segment of the first ratio behind the former and the second track segment of the second ratio before the latter in the adjacent placement tracks; wherein the shooting flip space includes: a spherical space with a radius length of the radius value between the first plane where the front, middle and last points on the first track segment are located and the second plane where the front, middle and last points on the second track segment are located, and the vertical distance from the center of the sphere to the first plane and the second plane does not exceed the distance threshold; in constraint three, the first ratio can be, for example, 30%; the second ratio can be, for example, 40%; the front, middle and last points refer to the starting point, midpoint and end point on the track segment; when the package is placed, it will move along the curve, so the plane where the front, middle and last points are located is unique; the radius value can be, for example, 45 cm; the distance threshold can be, for example, 55 cm;

[0129] Among them, the shooting trajectory planning constraints include:

[0130] Constraint 4: When the image acquisition device shoots along the i-th shooting trajectory in the shooting trajectory sequence, all standard shooting postures can be continuously changed to shoot the medical device package placed on the weighing platform along the i-th placement trajectory in the traversed local sequence; i = 1, 2, ..., N; N is the total number of shooting trajectories in the shooting trajectory sequence or the total number of placement trajectories in the traversed local sequence; i takes the same value each time; in constraint 4, all standard shooting postures are postures for shooting all sides of the package in mid-air;

[0131] And, constraint five, the end point of the i-th shooting trajectory in the shooting trajectory sequence is the center of the shooting flip space of the i-th group of adjacent placement trajectories in the traversed local sequence; i takes the same value each time; in constraint five, the image acquisition device can directly flip the lens as soon as possible after completing the shooting of a package, and flip the next package;

[0132] And, constraint six, when the image acquisition device shoots along the i-th and i+1-th shooting trajectories in the shooting trajectory sequence, at least one lens flip is performed in the shooting flip space of the i-th group of adjacent placement trajectories in the traversed local sequence; i takes the same value each time.

[0133] The image acquisition device in the embodiment of the present invention can be a mobile phone with a mobile arm, and the front and rear cameras of the mobile phone can be flipped. At the scene of medical device identification and counting, there will be a large number of packages piled up at the scene, leaving little space for the image acquisition device to move. In order to reduce the space occupied by movement, a mobile arm is used in combination with a mobile phone, and the mobile phone is equipped with a front camera and a rear camera. In the process of the mobile arm supporting the mobile phone to move and shoot, the mobile phone can increase the shooting range by flipping the lens (shooting switching between the front camera and the rear camera), and reduce the moving stroke of the mobile arm.

[0134] The present invention realizes intelligent planning of shooting trajectories according to the actual situation of placing trajectory sequences through local sequence division constraints and shooting trajectory planning constraints; the division of local sequences can effectively reduce the generation of invalid paths during the shooting process, avoid repeated movement of the image acquisition device, thereby saving time and improving shooting efficiency; in the medical device recognition and counting scenario, the site space is usually limited. By adopting the design of a mobile arm combined with a mobile phone and cooperating with the flip function of the lens, multi-directional coverage of shooting angles can be achieved, the physical moving space requirements of the equipment can be reduced, and the space utilization rate can be further improved, especially in a restricted environment, and the operation process can be significantly optimized; the planning of the shooting trajectory sequence takes into account the design of the shooting flipping space, the image acquisition device can quickly switch between different shooting trajectories, and can also ensure that different angles of each package can be fully photographed, and the shooting trajectory can flexibly respond to different packaging placement positions and directions; the basic navigation degree improves the adaptability of the equipment to complex environments and changing conditions, and increases the flexibility of the overall system.

[0135] The embodiment of the present invention provides a medical device automatic identification and counting system, such as Figure 3 As shown, including:

[0136] The acquisition module 1 is used to acquire the device image of each medical device through the image acquisition device when a large number of medical devices to be counted are placed on the weighing platform;

[0137] Identification module 2, used to identify the device image of each medical device and determine the device type of each medical device;

[0138] A determination module 3, for determining a counting result of each medical device based on a standard weight of a device type of each medical device and a weighing result of each medical device weighed by a weighing platform;

[0139] The output module 4 is used to output the counting results of each medical device.

[0140] The recognition module recognizes the device image of each medical device and determines the device type of each medical device, including:

[0141] Matching the device image of each medical device with a plurality of standard device images in a standard device image library;

[0142] When the matching is successful, the standard instrument types of the multiple standard instrument images that match the matching are used as the instrument type of each medical instrument.

[0143] The determination module determines the counting result of each medical device based on the standard weight of the device type of each medical device and the weighing result of each medical device weighed by the weighing platform, including:

[0144] Randomly combining the standard weights of device types of each medical device to obtain multiple weight sets;

[0145] Calculate the sum of the weights of each weight set;

[0146] The counting result of each medical device is determined based on the weight and the number of standard weights in the weight set with the smallest difference from the weighing result and the corresponding device type.

[0147] The acquisition module acquires the device image of each medical device through an image acquisition device, including:

[0148] If each medical device is a package, obtain the placement track of each medical device package placed on the weighing platform;

[0149] Based on the placement order of each medical device on the weighing platform, the placement trajectories corresponding to each medical device are sorted to obtain a placement trajectory sequence;

[0150] Based on the placement trajectory sequence, plan the shooting trajectory sequence;

[0151] Based on the shooting trajectory sequence, the image acquisition device is controlled to acquire the instrument image of each medical instrument.

[0152] The acquisition module plans a shooting trajectory sequence based on the placement trajectory sequence, including:

[0153] Based on the local sequence partitioning constraint, multiple local sequences are divided in the placement trajectory sequence;

[0154] Traverse each local sequence in turn according to the sequence order of each local sequence in the placement trajectory sequence;

[0155] During each traversal, a local shooting trajectory sequence of the traversed local sequence is planned based on the shooting trajectory planning constraints;

[0156] After traversing each local sequence, the local shooting trajectory sequence planned by each traversal is spliced ​​in the order of traversal to obtain a shooting trajectory sequence;

[0157] Among them, the local sequence partition constraints include:

[0158] Constraint 1: The same local sequence contains at least N consecutive placement trajectories in the placement trajectory sequence; N is a positive integer;

[0159] And,Constraint 2, the placement trajectories in different local sequences are different;

[0160] And, constraint three, in the same local sequence, there is a shooting flip space between the first track segment of the first proportion behind the former and the second track segment of the second proportion in front of the latter in the adjacent placement tracks; wherein the shooting flip space includes: a spherical space between the first plane where the front, middle and last three points on the first track segment are located and the second plane where the front, middle and last three points on the second track segment are located, with a radius length of the radius value, and the vertical distances from the center of the sphere to the first plane and the second plane respectively not exceeding the distance threshold;

[0161] Among them, the shooting trajectory planning constraints include:

[0162] Constraint 4: When the image acquisition device shoots along the i-th shooting trajectory in the shooting trajectory sequence, it can continuously change all standard shooting postures to shoot the medical device package placed on the weighing platform along the i-th placement trajectory in the traversed local sequence; i = 1, 2, ..., N; N is the total number of shooting trajectories in the shooting trajectory sequence or the total number of placement trajectories in the traversed local sequence; i takes the same value each time;

[0163] And, constraint 5: the end point of the i-th shooting trajectory in the shooting trajectory sequence is the center of the shooting flip space of the i-th group of adjacent placement trajectories in the traversed local sequence; i takes the same value each time;

[0164] And, constraint six, when the image acquisition device shoots along the i-th and i+1-th shooting trajectories in the shooting trajectory sequence, at least one lens flip is performed in the shooting flip space of the i-th group of adjacent placement trajectories in the traversed local sequence; i takes the same value each time.

[0165] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for automatically identifying and counting medical devices, characterized in that: include: When a large number of medical devices to be counted are placed on the weighing platform, the image of each medical device is collected by the image acquisition device; Identify the device image of each medical device and determine the device type of each medical device; Determine the counting result of each medical device based on the standard weight of the device type of each medical device and the weighing result of each medical device weighed by the weighing platform; Output the counting results of each medical device.

2. The method for automatic identification and counting of medical devices according to claim 1, characterized in that: The step of identifying the device image of each medical device and determining the device type of each medical device includes: Matching the device image of each medical device with a plurality of standard device images in a standard device image library; When the matching is successful, the standard instrument types of the multiple standard instrument images that match the matching are used as the instrument type of each medical instrument.

3. The method for automatic identification and counting of medical devices according to claim 1, characterized in that: The method of determining the counting result of each medical device based on the standard weight of the device type of each medical device and the weighing result of each medical device weighed by the weighing platform includes: Randomly combining the standard weights of device types of each medical device to obtain multiple weight sets; Calculate the sum of the weights of each weight set; The counting result of each medical device is determined based on the weight and the number of standard weights in the weight set with the smallest difference from the weighing result and the corresponding device type.

4. The method for automatic identification and counting of medical devices according to claim 1, characterized in that: The collecting of the device images of each medical device by the image collecting device includes: If each medical device is a package, obtain the placement track of each medical device package placed on the weighing platform; Based on the placement order of each medical device package on the weighing platform, the placement tracks of each medical device package are sorted to obtain a placement track sequence; Based on the placement trajectory sequence, plan the shooting trajectory sequence; Based on the shooting trajectory sequence, the image acquisition device is controlled to acquire the packaging image of each medical device package.

5. The method for automatic identification and counting of medical devices according to claim 4, characterized in that: The step of planning a shooting trajectory sequence based on the placement trajectory sequence includes: Based on the local sequence partitioning constraint, multiple local sequences are divided in the placement trajectory sequence; Traverse each local sequence in turn according to the sequence order of each local sequence in the placement trajectory sequence; During each traversal, a local shooting trajectory sequence of the traversed local sequence is planned based on the shooting trajectory planning constraints; After traversing each local sequence, the local shooting trajectory sequence planned by each traversal is spliced ​​in the order of traversal to obtain a shooting trajectory sequence; Among them, the local sequence partition constraints include: Constraint 1: The same local sequence contains at least N consecutive placement trajectories in the placement trajectory sequence; N is a positive integer; And,Constraint 2, the placement trajectories in different local sequences are different; And, constraint three, in the same local sequence, there is a shooting flip space between the first track segment of the first proportion behind the former and the second track segment of the second proportion in front of the latter in the adjacent placement tracks; wherein the shooting flip space includes: a spherical space between the first plane where the front, middle and last three points on the first track segment are located and the second plane where the front, middle and last three points on the second track segment are located, with a radius length of the radius value, and the vertical distances from the center of the sphere to the first plane and the second plane respectively not exceeding the distance threshold; Among them, the shooting trajectory planning constraints include: Constraint 4: When the image acquisition device shoots along the i-th shooting trajectory in the shooting trajectory sequence, it can continuously change all standard shooting postures to shoot the medical device package placed on the weighing platform along the i-th placement trajectory in the traversed local sequence; i = 1, 2, ..., N; N is the total number of shooting trajectories in the shooting trajectory sequence or the total number of placement trajectories in the traversed local sequence; i takes the same value each time; And, constraint 5: the end point of the i-th shooting trajectory in the shooting trajectory sequence is the center of the shooting flip space of the i-th group of adjacent placement trajectories in the traversed local sequence; i takes the same value each time; And, constraint six, when the image acquisition device shoots along the i-th and i+1-th shooting trajectories in the shooting trajectory sequence, at least one lens flip is performed in the shooting flip space of the i-th group of adjacent placement trajectories in the traversed local sequence; i takes the same value each time.

6. A medical device automatic identification and counting system, characterized in that: include: A collection module, used for collecting images of each medical device through an image collection device when a large number of medical devices to be counted are placed on the weighing platform; An identification module, used to identify the device image of each medical device and determine the device type of each medical device; A determination module, for determining a counting result of each medical device based on a standard weight of a device type of each medical device and a weighing result of each medical device weighed by a weighing platform; The output module is used to output the counting results of each medical device.

7. The medical device automatic identification and counting system according to claim 6, characterized in that: The recognition module recognizes the device image of each medical device and determines the device type of each medical device, including: Matching the device image of each medical device with a plurality of standard device images in a standard device image library; When the matching is successful, the standard instrument types of the multiple standard instrument images that match the matching are used as the instrument type of each medical instrument.

8. The medical device automatic identification and counting system according to claim 7, characterized in that: The determination module determines the counting result of each medical device based on the standard weight of the device type of each medical device and the weighing result of each medical device weighed by the weighing platform, including: Randomly combining the standard weights of device types of each medical device to obtain multiple weight sets; Calculate the sum of the weights of each weight set; The counting result of each medical device is determined based on the weight and the number of standard weights in the weight set with the smallest difference from the weighing result and the corresponding device type.

9. The medical device automatic identification and counting system according to claim 6, characterized in that: The acquisition module acquires the device image of each medical device through an image acquisition device, including: If each medical device is a package, obtain the placement track of each medical device package placed on the weighing platform; Based on the placement order of each medical device package on the weighing platform, the placement tracks of each medical device package are sorted to obtain a placement track sequence; Based on the placement trajectory sequence, plan the shooting trajectory sequence; Based on the shooting trajectory sequence, the image acquisition device is controlled to acquire the packaging image of each medical device package.

10. The medical device automatic identification and counting system according to claim 9, characterized in that: The acquisition module plans a shooting trajectory sequence based on the placement trajectory sequence, including: Based on the local sequence partitioning constraint, multiple local sequences are divided in the placement trajectory sequence; Traverse each local sequence in turn according to the sequence order of each local sequence in the placement trajectory sequence; During each traversal, a local shooting trajectory sequence of the traversed local sequence is planned based on the shooting trajectory planning constraints; After traversing each local sequence, the local shooting trajectory sequence planned by each traversal is spliced ​​in the order of traversal to obtain a shooting trajectory sequence; Among them, the local sequence partition constraints include: Constraint 1: The same local sequence contains at least N consecutive placement trajectories in the placement trajectory sequence; N is a positive integer; And,Constraint 2, the placement trajectories in different local sequences are different; And, constraint three, in the same local sequence, there is a shooting flip space between the first track segment of the first proportion behind the former and the second track segment of the second proportion in front of the latter in the adjacent placement tracks; wherein the shooting flip space includes: a spherical space between the first plane where the front, middle and last three points on the first track segment are located and the second plane where the front, middle and last three points on the second track segment are located, with a radius length of the radius value, and the vertical distances from the center of the sphere to the first plane and the second plane respectively not exceeding the distance threshold; Among them, the shooting trajectory planning constraints include: Constraint 4: When the image acquisition device shoots along the i-th shooting trajectory in the shooting trajectory sequence, it can continuously change all standard shooting postures to shoot the medical device package placed on the weighing platform along the i-th placement trajectory in the traversed local sequence; i = 1, 2, ..., N; N is the total number of shooting trajectories in the shooting trajectory sequence or the total number of placement trajectories in the traversed local sequence; i takes the same value each time; And, constraint 5: the end point of the i-th shooting trajectory in the shooting trajectory sequence is the center of the shooting flip space of the i-th group of adjacent placement trajectories in the traversed local sequence; i takes the same value each time; And, constraint six, when the image acquisition device shoots along the i-th and i+1-th shooting trajectories in the shooting trajectory sequence, at least one lens flip is performed in the shooting flip space of the i-th group of adjacent placement trajectories in the traversed local sequence; i takes the same value each time.

Images