Optionally, in the medical staff hand washing monitoring apparatus, the automatic monitoring of the hand washing process executed by the PS terminal of the processor system includes:

performing nearest neighbor matching learning on the hand washing image sequence and the hand key points based on the detected hand key points;

and finishing the learning of the similarity by using the hand washing image of the VGG based on the number of undetected or detected hand key points being less than the detection threshold.

Optionally, in the hand washing monitoring apparatus for medical staff, the programmable logic module further includes:

the optical flow method module is configured to perform foreground detection and extraction of hand actions in hand washing images, achieve the purpose of hand recognition detection, package the hand washing images into an IP core by using HLS, achieve the purpose at a hardware PL end, achieve a high-performance input/output streaming media interface by using an AXI-Stream interface, and prompt a CPU to control a synthesis module at an AXI-Lite interface; sharing data between the motion foreground detection result and a ZYNQ IP core of the processor system;

and the HDMI-OUT module is configured to receive the identification result of the hand washing monitoring equipment, is connected with the large-screen display terminal and completes visualization of the result.

Optionally, in the hand washing monitoring apparatus for medical staff, the data transmission between the PS terminal of the processor system and the programmable logic module PL is implemented by using an AXI bus to complete the transmission of high performance data, so as to meet the real-time requirement of the hand washing monitoring apparatus for medical staff on image processing; and the access of the PL end hardware IP core to the system memory DDR of the read-write PS end is finished by utilizing the AXI _ HP interface.

Optionally, in the hand washing monitoring apparatus for medical staff, the system hardware path design of the hand washing monitoring apparatus for medical staff includes:

after the programmable logic module PL finishes the design of a hardware acceleration IP core, the specific function of the IP core is described through configuration parameters; the parameters comprise the width and height of an input image and a data type;

setting interface constraint, parameter constraint and function name constraint of the IP core by using a Vivado HLS comprehensive tool;

the interface constraint sets the ARM memory as a slave device, and the hardware IP core is a master device M _ AXI interface, so that the IP core actively reads ARM memory data;

parameter constraint, namely setting an ARM end as a master device, an IP core as a slave device, and setting an interface as S _ AXI _ Lite to realize that ARM actively configures parameters of the hardware IP core;

function name constraint, setting an ARM end as a master device and a hardware IP core as a slave device, and realizing that the ARM actively controls the start and stop of the hardware IP core;

the method comprises the steps of packaging an IP core, building a ZYNQ-based SOC platform, and completing the connection of hardware paths among the ZYNQ processing system IP core, a color space conversion IP core, an optical flow method IP core, an AXI interconnection device and a clock reset module;

designing a path of a hardware PL end by using Block Design of a Vivado platform through AXI VDMA;

the PS end of the processor system sends an instruction to the AXI VDMA through the AXI-Lite interface, the AXI VDMA completes data exchange with the cache DDR memory through the AXI _ HP interface and caches the data in an internal register;

the hardware acceleration IP core utilizes the AXI-Stream bus to complete the read-write operation of data in the AXI VDMA.

Optionally, the hand washing monitoring device for medical staff further comprises:

a screen display module configured to display a hand washing detection result;

the screen display module is configured with a Video Timing Controller IP core and a Video data Stream conversion AXI-Stream IP core to generate a Video time sequence signal, and control of factual data Stream in the hand washing monitoring device of medical staff is completed;

because the data stream protocol and the data interface type input or output by the HDMI-OUT module are not consistent with the system VDMA terminal, when the hand washing result is displayed, a hardware acceleration IP core is designed and configured, the conversion of the hand washing image stream media format is completed, the conversion of the video data stream protocol and the conversion of the interface data type are included, and the hand washing detection result is displayed.

The invention also provides a hand washing monitoring method for medical staff, which comprises the following steps:

the image acquisition module acquires hand washing images and provides the hand washing images to other systems and modules;

the PL end of the programmable logic module acquires a hand washing image, and gray scale conversion and key motion frame extraction of the hand washing image are realized through color space conversion and a motion foreground detection optical flow algorithm, so that hand detection and tracking are realized; and

the PS end of the processor system runs a hand key point acquisition and hand washing process automatic monitoring algorithm through an ARM; and the detection of hand washing behavior is finished by calling the hardware IP core through the written main control code.

The programmable logic module acquires a hand washing image, and performs image preprocessing on the hand washing image through a hand washing image algorithm to obtain a preprocessing result; and

the processor system runs the hand key points through the ARM to obtain and automatically monitor the algorithm in the hand washing process, and writes the main control code to complete the calling of the operation IP core. And finishing the hand washing image learning and monitoring. And finishing the hand washing image learning and monitoring according to the detection result.

And at the programmable logic PL terminal, completing the identification of human hands, including gray scale conversion in color space conversion and an optical flow method for detecting a moving foreground. Designing a graying conversion and foreground detection module by using a Vivado HLS tool, packaging the graying conversion and foreground detection module into an IP core, and building an SOC platform based on ZYNQ;

at the PS end of the processor system, the acquisition of key points of hands and the automatic monitoring of the hand washing process are completed; and the calling of the operation IP core is completed by writing the master control code. And data interaction between the software PS and the hardware PL is carried out through the AXI VDMA by utilizing a corresponding API function, and the cooperative work of the software and the hardware is fully realized.

The processor system includes:

the hand key point acquisition module is configured to detect and extract extracted hand key points in the hand washing image based on a deep learning framework, so as to realize the detection of the hand key points;

the nearest neighbor matching learning module is configured to analyze the hand gesture of the person, the hand action of the person, the matching degree of the hand gesture of the person and the standard flow consistency and the matching degree of the hand action of the person and the standard flow consistency according to the hand key points after the hand key points are detected by the hand washing process automatic monitoring module; and

however, in an actual scene, since the hands of a person are seriously overlapped and shielded during hand washing, the difficulty of detecting key points of the hands is increased. The hand washing image similar learning module based on the VGG is configured to extract the hand posture key point features by adopting a multilayer convolution and Relu layers when the hand positioning and posture acquisition module does not detect the hand key points, and finally, one-dimensional probability information of similarity is obtained by using a full connection layer, and an original value range is mapped to a value range space of the probability so as to carry out the hand washing image similar learning of the VGG.

Optionally, in the hand washing monitoring device for medical staff, the hand posture acquiring module includes:

and the hand key point extraction module is configured to detect and extract the hand key points in the hand washing image.

The hand washing method based on the hand recognition and monitoring in the hand washing image and the detection and extraction of the key points of the hand in the hand washing image. In the algorithm implementation process, a lot of multiplication and floating point operations are included, and the processing speed of the algorithm is reduced due to the limited internal storage resources of the programmable logic module. Thus, the partial algorithm is processed in the processor system.

Optionally, in the hand washing monitoring apparatus for medical staff, the processor system runs a matching learning algorithm of the hand washing action through an ARM, including:

Because the algorithm for acquiring the key points of the hands and automatically monitoring the hand washing process contains a large amount of multiplication and floating point operation, and the programmable logic module does not contain a hardware operation unit required by the floating point operation, if the floating point operation is carried out, a large amount of internal logic resources are occupied, and floating point fixed-point operation is required to be adopted for processing the floating point operation, so that the processing speed of the algorithm is reduced.

Thus, a matching learning algorithm for hand washing actions is implemented in the processor system.

the optical flow method module is configured to perform foreground detection and extraction of human hand actions in the hand washing image to obtain a motion foreground detection result, and the motion foreground detection result is shared with the HDMI-OUT module and/or the processor system;

and the HDMI-OUT module is configured to receive the motion foreground detection result and perform rear-end detection and identification.

Because the back end detection and identification are involved, the time and the effect of the foreground detection and extraction of the human hand action have great influence on the whole system, and the optical flow method adopted by the foreground detection and extraction of the human hand action is put on the programmable logic module for realization.

Optionally, in the medical staff hand washing monitoring apparatus, the data transmission between the processor system and the programmable logic module uses an AXI bus to complete the transmission of high performance data, so as to meet the real-time requirement of the medical staff hand washing monitoring apparatus on image processing.

Optionally, in the hand washing monitoring device for medical staff, the system hardware path design of the hand washing monitoring device for medical staff includes:

after the programmable logic module finishes the design of a hardware acceleration IP core, the connection of a hardware channel is carried out, and the realization of the hardware channel is finished by utilizing VivadoHLS and AXI VDMA;

the processor system sends an instruction to the AXI VDMA through the AXI-Lite interface, the AXI VDMA completes data exchange with the cache and the DDR memory and caches in an internal register;

the hardware acceleration IP core completes the read-write operation of data in the AXI VDMA by utilizing an AXI-Stream port.

a screen display module configured to display a hand washing detection result;

because the data type output by the HDMI-OUT module is inconsistent with the data type of the system IP end, when the hand washing result is displayed, a hardware acceleration IP core is designed and configured to complete the conversion of the hand washing image streaming media format and complete the display of the hand washing detection result.

The inventor of the invention finds that due to the working property, the workload of medical staff is large, only the conventional treatment can be considered, and the hand washing before each operation is difficult to realize in the actual work. The degree of heavy work is a major factor affecting hand washing compliance of medical personnel.

In the medical staff hand washing monitoring device provided by the invention, hand washing images are collected through the image collection module, the programmable logic module identifies and detects hands, and the processor system runs the automatic monitoring algorithm of the hand washing process and the acquisition of key points of both hands through the ARM; the main control code written by the ARM end realizes the calling of the IP core of the programmable logic module, realizes the cooperative cooperation of software and hardware, completes the image learning and monitoring of hand washing, can intelligently collect and analyze the hand washing process based on the artificial intelligence combined with the embedded technology, effectively identifies the unqualified steps in the hand washing process, effectively warns, and is an important means for standardizing the hand washing of medical personnel.

Because the medical staff hand washing monitoring system has relatively high requirements on the aspects of management accuracy, real-time performance, power consumption and the like, and a single platform is difficult to meet the requirements, the extensible embedded platform ARM + FPGA architecture based on Xilinx corporation enables the system development process to utilize the characteristics of strong parallelism and rich logic resources of the FPGA and the advantages of ARM building operation systems. In conclusion, the invention has the following beneficial effects:

(1) the performance of the heterogeneous architecture-based system is 2.4 times of that of a system based on a single PS architecture, the detection rate of the system can be accelerated by using a heterogeneous platform, the detection time is shortened, and the requirement on the real-time performance of the system can be met.

(2) The invention is based on the architecture of an embedded platform ARM + FPGA, and can reduce the working hours required by visual inspection, supervision, confirmation and manual recording to 0. The invention also has the educational effect, and prompts each medical worker to master the correct hand washing method, thereby realizing the standardization of the hand washing mode.

(3) Aiming at the standard hand hygiene of medical staff, the invention adopts computer vision, artificial intelligence technology and embedded technology, and can complete the standard AI quantification standard.

Drawings

FIG. 1 is a diagram of a software and hardware system of a hand washing monitoring device for medical personnel in accordance with an embodiment of the present invention;

fig. 2 is a schematic diagram of a software and hardware system of a hand washing monitoring device for medical personnel according to another embodiment of the present invention.

Detailed Description

The invention is further elucidated with reference to the drawings in conjunction with the detailed description.

It should be noted that the components in the figures may be exaggerated and not necessarily to scale for illustrative purposes. In the figures, identical or functionally identical components are provided with the same reference symbols.

In the present invention, "disposed on …", "disposed over …" and "disposed over …" do not exclude the presence of an intermediate therebetween, unless specifically indicated otherwise. Further, "disposed on or above …" merely indicates the relative positional relationship between two components, and may also be converted to "disposed below or below …" and vice versa in certain cases, such as after reversing the product direction.

In the present invention, the embodiments are only intended to illustrate the aspects of the present invention, and should not be construed as limiting.

In the present invention, the terms "a" and "an" do not exclude the presence of a plurality of elements, unless otherwise specified.

It is further noted herein that in embodiments of the present invention, only a portion of the components or assemblies may be shown for clarity and simplicity, but those of ordinary skill in the art will appreciate that, given the teachings of the present invention, required components or assemblies may be added as needed in a particular scenario. In addition, features in different embodiments of the invention may be combined with each other, unless otherwise specified. For example, a feature of the second embodiment may be substituted for a corresponding or functionally equivalent or similar feature of the first embodiment, and the resulting embodiments are likewise within the scope of the disclosure or recitation of the present application.

It is also to be noted here that, within the scope of the present invention, the expressions "identical", "equal" and the like do not mean that the two values are absolutely equal, but allow a certain reasonable error, that is, the expressions also cover "substantially identical", "substantially equal". By analogy, in the present invention, the terms "perpendicular", "parallel" and the like in the directions of the tables also cover the meanings of "substantially perpendicular", "substantially parallel".

The numbering of the steps of the methods of the present invention does not limit the order in which the method steps are performed. Unless specifically stated, the method steps may be performed in a different order.

The hand washing monitoring device for medical staff, which is proposed by the present invention, is further described in detail with reference to the accompanying drawings and specific embodiments. The advantages and features of the present invention will become more apparent from the following description. It is to be noted that the drawings are in a very simplified form and are not to precise scale, which is merely for the purpose of facilitating and distinctly claiming the embodiments of the present invention.

The invention aims to provide a hand washing monitoring device for medical staff, which aims to solve the problem of poor hand washing compliance of the existing medical staff.

In order to achieve the above object, the present invention provides a hand washing monitoring device for medical staff, comprising: an image capture module configured to capture hand washing images and provide them to other systems and modules; the programmable logic module is configured to identify and detect hands, and identify and detect hand images moving during hand washing through a hand washing image algorithm; the processor system is configured to complete the acquisition of key points of hands and the automatic monitoring of the hand washing process through ARM; and meanwhile, calling the programmable logic module by writing the main control code. The invention provides a hand washing monitoring device for medical staff, which comprises the following components as shown in figure 1: an image capture module configured to capture hand washing images and provide them to other systems and modules; the programmable logic module is configured to acquire a hand image, perform color space conversion and motion foreground optical flow method detection on the hand image through a hand washing image algorithm, and acquire the hand image from a motion image; and a processor system configured to complete the acquisition of key points of hands and the automatic monitoring of the hand washing process through ARM; and meanwhile, calling the programmable logic module by writing the main control code.

In an embodiment of the present invention, in the medical staff hand washing monitoring apparatus, as shown in fig. 1, the image acquisition module adopts a high-definition camera to complete hand washing image acquisition; the image acquisition module reads the acquired hand washing image from the cache, stores the hand washing image in the DDR memory, and processes the acquired hand washing image according to the first address of the DDR memory; the programmable logic module processes the color space conversion and the motion foreground detection optical flow method in the programmable logic module; the programmable logic module adopts a packaged hardware acceleration IP core, and the hardware acceleration IP core adopts Vivado HLS for design and packaging. Image preprocessing: the color space conversion with a large amount of calculation is processed at the PL end of the system. And adopting an encapsulated hardware acceleration IP core at the PL end of the hardware. And the hardware IP is designed and packaged by adopting Vivado HLS.

In one embodiment of the present invention, in the hand washing monitoring apparatus for medical staff, as shown in fig. 1, the processor system comprises: the hand key point acquisition module is configured to be a deep learning-based framework to realize extraction of hand key points, and in one embodiment of the invention, the hand key point acquisition module can also be used for positioning human hand gestures and analyzing whether the human hand gestures meet a standard flow to realize human hand recognition and monitoring; the nearest neighbor matching learning module is configured to analyze the human hand gesture, analyze the human hand action, analyze the matching degree of the human hand gesture and the standard process consistency and analyze the matching degree of the human hand action and the standard process consistency according to the hand key points after the hand key points are detected by the hand key point acquisition module; and the washing image similar learning module based on the VGG is configured to analyze the human hand motions in various forms to obtain a first analysis result, analyze the postures of the human hands in the standard flow to obtain a second analysis result and compare and match the first analysis result and the second analysis result when the hand key point acquisition module does not detect the hand key points so as to perform washing image similar learning of the VGG.

Acquiring key points of the hands: aiming at 6 actions of standard hand washing, preliminarily detecting key points through multi-view guidance based on a deep learning framework, constructing triangles on the key points according to the pose of a camera to obtain key 3D positions, re-projecting the obtained 3D positions to 2D images at different views, detecting and extracting extracted hand key points in the hand washing images by using the 2D images and key point labeling data sets, and realizing hand key point detection; in one embodiment of the invention, it is also possible to implement: and positioning and analyzing whether the gesture of the human hand meets the standard, and mainly analyzing the gesture and the action of the human hand and matching the gesture and the action with the consistency of a standard flow. Analyzing the hand motions in various forms and analyzing the postures of the hand in the standard operation flow. The process is mainly divided into two processes: 1. identifying and monitoring human hands based on YOLO; 2. and detecting and extracting key points of the human hand. The process algorithm implementation process comprises a plurality of multiplications and floating point operations, and the processing speed of the algorithm is reduced due to the limited internal storage resources of the PL end of the hardware. Therefore, this part is handled in the software PS part.

In an embodiment of the present invention, in the hand washing monitoring device for medical staff, as shown in fig. 1, the hand positioning and posture acquiring module includes: a hand recognition and monitoring module configured to recognize and monitor a hand in the hand washing image based on YOLO; and a hand key point extraction module configured to detect and extract hand key points in the hand washing image. The method is characterized in that the method is based on YOLO to identify and monitor hands in the hand washing image and detect and extract key points of the hands in the hand washing image, a lot of multiplications and floating point operations are included in the algorithm implementation process, and the processing speed of the algorithm is slowed down due to limited internal storage resources of the programmable logic module. Thus, the partial algorithm is processed in the processor system.

Automatic monitoring of the hand washing process: based on the detected key points of the hand, matching and learning the image sequence and the key points of the hand; and based on undetected or difficultly detected hand key points, learning of similarity is completed by adopting a VGG-loss algorithm. Because the algorithm in the process also contains a large number of multiplications and floating-point operations, and because the PL end does not contain a hardware operation unit required by the floating-point operations, if the floating-point operations are carried out, a large number of internal logic resources are occupied, and floating-point fixed-point operation is required to be adopted for processing the floating-point operations, so that the processing speed of the algorithm is reduced. Thus, the process is implemented on the software PS side.

In an embodiment of the present invention, in the hand washing monitoring apparatus for medical staff described in the above, as shown in fig. 1, the processor system runs an automatic hand washing process monitoring algorithm through an ARM, and the automatic hand washing process monitoring algorithm includes: based on the detected hand key points, performing nearest neighbor matching learning on the hand washing image sequence and the hand key points by using a nearest neighbor based matching learning algorithm; and finishing the learning of the similarity by using the hand washing image of the VGG based on the number of undetected or detected hand key points being less than the detection threshold. Because the matching learning algorithm of the hand washing action also comprises a large number of multiplications and floating point operations, and because the programmable logic module does not comprise a hardware operation unit required by the floating point operation, if the floating point operation is carried out, a large number of internal logic resources are occupied, and the floating point fixed-point operation is required to be adopted for processing the floating point operation, so that the processing speed of the algorithm is reduced. Thus, a matching learning algorithm for hand washing actions is implemented in the processor system.

In an embodiment of the present invention, in the hand washing monitoring apparatus and/or method for medical staff, as shown in fig. 1, the programmable logic module further includes: the optical flow method module is configured to perform foreground detection and extraction of human hand actions in the hand washing image, obtain a motion foreground detection result, and share the motion foreground detection result with a storage part in the AXI VDMA; and the HDMI-OUT module is configured to receive the hand washing detection result of the medical staff and perform back-end display. In one embodiment of the invention, the method can also be used for realizing the foreground detection and extraction of the human hand motion by putting an optical flow method adopted by the foreground detection and extraction of the human hand motion on a programmable logic module, because the time and the effect of the foreground detection and extraction of the human hand motion have great influence on the whole system, relating to the back-end detection and identification.

The human hand recognition detection module: the results of color space conversion and motion foreground detection can be acquired by the key points of the back-end hand and shared by the automatic hand washing process monitoring module, so that the time and effect of foreground detection and extraction have great influence on the whole system, and the color space conversion and the hand washing motion foreground detection algorithm optical flow method is put at the PL end of hardware for realization.

In an embodiment of the medical staff hand washing monitoring apparatus and/or method, as shown in fig. 1, the data transmission between the processor system and the programmable logic module uses an AXI bus to complete the transmission of high performance data, so as to meet the real-time requirement of the medical staff hand washing monitoring apparatus on image processing. Because the real-time performance of the system on image processing is higher, the data transmission of the PS end and the PL end adopts the AXI bus to complete the transmission of high-performance data.

In one embodiment of the present invention, in the hand washing monitoring apparatus for medical staff, as shown in fig. 1, the system hardware path design of the hand washing monitoring apparatus for medical staff includes: after the programmable logic module finishes the design of a hardware acceleration IP core, the connection of a hardware channel is carried out, and the realization of the hardware channel is finished by utilizing VivadoHLS and AXI VDMA; the processor system sends an instruction to the AXI VDMA through the AXI-Lite interface, the AXI VDMA completes data exchange with the cache and the DDR memory and caches in an internal register; the hardware acceleration IP core utilizes the AXI-Stream port to complete the read-write operation of data in the AXI VDMA.

The hardware path of the system is realized as follows: after the PL end of the system completes the IP core design, the connection of a hardware access is needed, and the realization of the hardware access is completed by using Vivado and AXI VDMA. The PS end sends an instruction to the AXI VDMA through the AXI-Lite interface, the VDMA and the cache DDR complete data exchange, and the cache and the internal register are connected; and then the IP core completes the read-write operation of the data in the VDMA by using the AXI-Stream port.

In an embodiment of the present invention, the hand washing monitoring device for medical staff, as shown in fig. 1, further includes: the screen display module is configured to display the hand washing detection result of the medical staff; the screen display module is configured with a Video Timing Controller IP core and a Video data Stream conversion AXI-Stream IP core to generate a Video time sequence signal, and the control of real-time data Stream in the hand washing monitoring device and/or method of medical staff is completed; because the data type input or output by the HDMI-OUT module is not consistent with the data type of the system IP end, when the hand washing result is displayed, a hardware acceleration IP core is designed and configured, the conversion of the hand washing image streaming media format is completed, and the display of the hand washing detection result is completed.

And designing and configuring a Video Timing Controller IP core to be used together with a Video data Stream conversion AXI-Stream IP core to generate a Video time sequence signal, and finishing the control of the fact data Stream in the system. The data type input or output by the HDMI terminal is not consistent with the data type of the system IP terminal, and when the hand washing result is displayed, a hardware IP needs to be designed and configured to complete conversion of the hand washing image streaming media format and display of the hand washing detection result.

The invention also provides a hand washing monitoring method for medical staff, which comprises the following steps: the image acquisition module acquires hand washing images and provides the hand washing images to other systems and modules; the programmable logic module acquires a first hand image, and performs image color space conversion and motion foreground optical flow method detection on the hand image through a hand washing image algorithm to obtain a second hand image; and the processor system runs a hand key point through the ARM to acquire and automatically monitor an algorithm in the hand washing process, writes a main control code and completes the calling of the operation IP core. And finishing the hand washing image learning and monitoring. And finishing the learning and monitoring of the hand washing image according to the detection result.

In the medical staff hand washing monitoring device provided by the invention, the hand washing image is collected through the image collection module, the programmable logic module identifies and detects the hand in the hand washing image, the processor system runs the hand key point through the ARM processor system to obtain and wash the hand process automatic monitoring algorithm, and writes the main control code to complete the calling of the operation IP core. The hand washing image learning and monitoring can be completed, the hand washing process can be intelligently collected and analyzed based on an artificial intelligence technology, unqualified steps in the hand washing process can be effectively identified, effective warning is carried out, and the method is an important means for standardizing hand washing of medical workers.

Common DSP, ARM, PowerPC, MIPS, FPGA and the like of a traditional embedded integrated circuit application level chip, and the FPGA has the advantages of good flexibility, rich resources and high (parallel) speed of repeated programming (programmability). In the past application, an ARM is often used as a main control in a scene, an FPGA is externally hung on a peripheral parallel RAM (random access memory) bus of the ARM, and the FPGA is used for constructing high-speed data acquisition or operation, which is common in the industries of communication, medical electronics and the like;

currently, the related technologies such as artificial intelligence AI and machine learning ML have been widely applied and have been greatly successful in various industries. However, most of the functions are realized by using a PC, and with the development of embedded systems such as ARM, DSP, FPGA, and the like, the features of low power consumption and strong computing performance are increasingly applied to machine learning. Most designers use a single platform for pedestrian detection, and although the system performance is improved, the degree of improvement is limited.

The inventor of the invention discovers through research that the software and hardware system of the medical staff hand washing monitoring device has relatively high requirements on management accuracy, real-time performance, power consumption and the like, and a single platform is difficult to meet the requirements, so the inventor overcomes the technical bias in the prior art, and adopts an extensible embedded platform ARM + FPGA architecture based on Xilinx company to monitor the hand washing of the medical staff, so that the characteristics of strong parallelism and rich logic resources of the FPGA can be utilized in the system development process, and the advantage of an operation system can also be built by utilizing the ARM.

As shown in fig. 2, Xilinx is taken as a leader of an FPGA (field Programmable Gate array), and a dedicated CPU (central processing unit) hardmac and the FPGA are integrated into one chip, so as to generate a brand new heterogeneous platform, which is called an All Programmable system on a chip (All Programmable SoC); the processing system is PS, the programmable logic is PL, the PS processing system comprises an ARM hard core, hardware can be developed by utilizing python language, a perfect OpenCV library is supported, a PL end can customize needed operation or peripheral equipment, and the requirements of software and hardware systems on management accuracy, instantaneity, power consumption and the like are met perfectly through communication between an AXI bus and the PS. As shown in fig. 2, one ZYNQ is used to implement the design architecture of the conventional ARM external FPGA, so that the total cost is reduced, the performance is improved, the power consumption is reduced, the size is reduced, and the reliability is improved; the birth of the brand new platform plays a great role in promoting the development of global information technology. On one hand, the design structure of the embedded system is more flexible, the volume is obviously reduced, and the reliability and the overall performance of the system are obviously improved; on the other hand, the FPGA can enter the application field of an embedded system, the application range of the FPGA is greatly expanded, the advantages are integrated, the scheme of the invention can realize the parallel schemes in multiple aspects such as image graying processing on the hand washing image, recognition and detection on a moving hand, key point detection and extraction on the hand, nearest neighbor based matching learning algorithm, learning of VGG-loss hand washing image completion similarity and the like by a moving foreground detection optical flow method, can complete redundancy of various monitoring means to a higher degree, can perform preprocessing, detection and extraction, and respectively adopt different modes to process under two conditions that key points are extracted and key points are not extracted.

In conclusion, the invention has the following beneficial effects:

(2) The invention is based on the architecture of an embedded platform ARM + FPGA, and can reduce the working hours required by visual inspection, supervision and confirmation and manual recording to 0. The invention also has the educational effect, and prompts each medical worker to master the correct hand washing method, thereby realizing the standardization of the hand washing mode.

In summary, the above embodiments describe the various configurations of the hand washing monitoring device for medical staff in detail, and it is understood that the present invention includes, but is not limited to, the configurations listed in the above embodiments, and any modifications based on the configurations provided by the above embodiments are within the scope of the present invention. One skilled in the art can take the contents of the above embodiments to take a counter-measure.

In the present specification, the embodiments are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. For the system disclosed by the embodiment, the description is relatively simple because the system corresponds to the method disclosed by the embodiment, and the relevant points can be referred to the method part for description.

The above description is only for the purpose of describing the preferred embodiments of the present invention, and is not intended to limit the scope of the present invention, and any variations and modifications made by those skilled in the art based on the above disclosure are within the scope of the appended claims.

Claims

1. A hand washing monitoring device for medical personnel, comprising:

an image acquisition module configured to acquire a hand washing image;

the programmable logic module is configured to identify and detect human hands, and performs image color space conversion and motion foreground optical flow method processing on hand washing images through a hand washing image algorithm to finish detection and tracking of the human hands; and

2. The medical staff hand washing monitoring device of claim 1, wherein the image acquisition module uses a high-definition camera to complete the acquisition of hand washing images;

the image acquisition module reads the acquired hand washing image from the cache, stores the hand washing image in the DDR memory, and processes the acquired hand washing image according to the first address of the DDR memory;

3. Medical personnel hand washing monitoring device as claimed in claim 2, wherein the processor system comprises:

wherein, e in the formula_iIs the original logical value.

4. Medical personnel hand washing monitoring device as claimed in claim 3, wherein the hand washing process automatic monitoring by the processor system PS running a matching learning algorithm for hand washing actions comprises:

5. The medical personnel hand washing monitoring device of claim 4, wherein the programmable logic module further comprises:

6. The hand washing monitoring device for medical staff as claimed in claim 1, wherein the data transmission between the PS terminal of the processor system and the PL is performed by an AXI bus to complete the transmission of high performance data, so as to meet the real-time requirement of the hand washing monitoring device for medical staff on image processing; and the access of the PL end hardware IP core to the system memory DDR of the read-write PS end is finished by utilizing the AXI _ HP interface.

7. The medical personnel hand washing monitoring device of claim 2, wherein the system hardware path design of the medical personnel hand washing monitoring device comprises:

the PS end of the processor system sends an instruction to the AXI VDMA through an AXI-Lite interface, the AXI VDMA completes data exchange with the cache DDR memory through an AXI-HP interface and caches in an internal register;

8. The medical personnel hand washing monitoring device of claim 6, further comprising:

a screen display module configured to display a hand washing detection result;

9. A hand washing monitoring method for medical personnel, comprising: