Intelligent examination data processing method

Abstract

The application provides a kind of intelligent examination data processing method, by receiving the target enterprise's workstation template map configured by management end, workstation template map includes multiple identification areas, identification area has corresponding examination end;Receive the examination data uploaded by each examination end, based on the examination data, the corresponding identification area in the workstation template map is handled, and the theoretical examination map is obtained;Obtain the workstation image collected by acquisition device, based on examination identification strategy and workstation template map, the workstation image is identified, and the update data of the examination end is obtained, and based on update data, the actual examination map is updated to the corresponding identification area;Based on video frame selection strategy, multiple workstation images in a predetermined time period are selected and processed, and the examination backtracking video of each examination end is obtained, and the actual examination map and the examination backtracking video are sent to the management end.

Description

Intelligent assessment data processing methods Technical Field

[0001] This invention relates to data processing technology, and more particularly to an intelligent assessment data processing method. Background Technology

[0002] Intelligent performance evaluation data processing is the process by which enterprises use computer hardware and software technologies to effectively collect, store, process, and apply data. Enterprise data can be diverse, such as the management of attendance and clock-in data within the enterprise.

[0003] In existing technologies, companies typically use electronic software for attendance tracking, automatically generating attendance sheets based on the data. These sheets display employee attendance information by showing the time they clocked in. However, this method cannot accurately reflect actual employee attendance. Furthermore, errors in attendance tracking can occur, such as when an employee clocks in within the designated time but leaves the office after clocking in. In such cases, the company cannot review the data, resulting in a lack of accurate attendance information.

[0004] Therefore, how to visualize enterprise data to facilitate viewing and ensure data accuracy has become an urgent problem to be solved. Summary of the Invention

[0005] This invention provides an intelligent performance evaluation data processing method that can record and display the actual attendance information of enterprise employees in the form of images, while retaining relevant data information to create performance evaluation retrospective videos for feedback and review, thereby improving the intuitiveness and authenticity of attendance information.

[0006] A first aspect of this invention provides an intelligent assessment data processing method, comprising:

[0007] The system receives a workstation template image of the target enterprise configured by the management terminal. The workstation template image includes multiple identification areas, and each identification area has a corresponding assessment terminal.

[0008] Receive assessment data uploaded by each of the assessment terminals, and process the corresponding identification area in the workstation template diagram based on the assessment data to obtain a theoretical assessment diagram;

[0009] The system acquires workstation images collected by the acquisition device, identifies the workstation images based on the assessment recognition strategy and workstation template, obtains updated data for the assessment end, and updates the corresponding recognition areas in the theoretical assessment image based on the updated data to obtain the actual assessment image.

[0010] Based on a video frame selection strategy, multiple workstation images within a preset time period are selected and processed to obtain assessment retrospective videos for each assessment end. The actual assessment image and the assessment retrospective video are then sent to the management end.

[0011] Optionally, in one possible implementation of the first aspect, receiving the assessment data uploaded by each of the assessment terminals, and processing the corresponding identification area in the workstation template diagram based on the assessment data to obtain a theoretical assessment diagram, includes:

[0012] Receive assessment data uploaded by each of the assessment terminals, the assessment data including attendance data and absence data, determine a first pixel value based on the attendance data, and determine a second pixel value based on the absence data;

[0013] Based on the assessment end, the first pixel value or the second pixel value is used to update the border of the corresponding recognition area in the workstation template image to obtain the theoretical assessment image.

[0014] Optionally, in one possible implementation of the first aspect, the acquisition of the workstation image acquired by the acquisition device, the recognition of the workstation image based on the assessment recognition strategy and the workstation template image to obtain updated data of the assessment end, and the updating of the corresponding recognition area in the theoretical assessment image based on the updated data to obtain the actual assessment image, includes:

[0015] The identification template corresponding to the acquisition device is retrieved from the workstation template image, the identification area in the identification template is used as the judgment area, and the workstation image collected by the acquisition device is obtained.

[0016] The recognition template is superimposed on the workstation image, and personnel recognition is performed on the image of the judgment area in the recognition template to obtain updated data. Based on the updated data, the corresponding recognition area in the theoretical assessment image is updated to obtain the actual assessment image.

[0017] Optionally, in one possible implementation of the first aspect, the step of performing personnel identification on the image of the judgment area in the recognition template to obtain updated data, and updating the corresponding recognition area in the theoretical assessment image based on the updated data to obtain the actual assessment image, includes:

[0018] The number of workstation images of personnel at the assessment end corresponding to each judgment area in the recognition template is used as the assessment judgment quantity.

[0019] If the number of assessment judgments is greater than or equal to the preset number, then the current data of the corresponding assessment terminal is determined to be attendance data.

[0020] If the number of assessment judgments is less than the preset number, then the current data of the corresponding assessment terminal is determined to be absence data;

[0021] If the current data of the assessment terminal is inconsistent with the assessment data, the third pixel value is retrieved as the update data, and the corresponding recognition area of ​​the assessment terminal in the theoretical assessment diagram is used as the update area.

[0022] The border of the updated area is updated based on the updated data to obtain the actual assessment image.

[0023] Optionally, in one possible implementation of the first aspect, it also includes:

[0024] Retrieve the job attributes from the judgment area of ​​the recognition template, whereby the job attributes include R&D attributes and communication attributes;

[0025] Based on the aforementioned R&D attributes, the corresponding judgment area is designated as the R&D area, and based on the communication attributes, the corresponding judgment area is designated as the communication area.

[0026] The continuity of focus in the R&D areas is assessed to obtain the R&D focus level of each R&D area;

[0027] Intermittent attention levels are assessed in the communication areas to obtain the communication attention level for each area.

[0028] Optionally, in one possible implementation of the first aspect, the step of performing a continuous focus determination on the R&D areas to obtain the R&D focus of each R&D area includes:

[0029] The orientation of the display devices in each R&D area is retrieved, and the facial orientation of the personnel in the R&D area is identified. The time period when the device orientation is opposite to the facial orientation is recorded as the first conditional time period for each R&D area.

[0030] The time periods during which personnel in each R&D area trigger input devices are used as the second conditional time periods for each R&D area.

[0031] Based on the intersection of the first conditional time period and the second conditional time period of each R&D zone, the conditional superposition time period of each R&D zone is obtained;

[0032] The image at any time during the time period corresponding to the conditions overlay in each R&D area is selected as the reference image, and the images at two times adjacent to the reference image are used as comparison images;

[0033] Pixels at the same position in the reference image and the comparison image are selected as the origin of coordinates, and the reference image and the comparison image are processed by coordinate transformation based on the origin of coordinates;

[0034] Pixels with different pixel values ​​in the baseline image and the comparison image under the same coordinates are identified as changed pixels. The number of changes of the changed pixels and the total number of pixels in the baseline image are obtained.

[0035] The change percentage is obtained based on the ratio of the number of changes to the total number. When the change percentage is less than the preset change percentage, the moment of the corresponding benchmark image is taken as the R&D focus moment.

[0036] The R&D focus time is obtained by statistically analyzing the R&D focus time. Based on the ratio of the R&D focus time to the preset working time, the R&D focus level of each R&D area is obtained.

[0037] Optionally, in one possible implementation of the first aspect, the step of intermittently judging the focus of the communication areas to obtain the communication focus of each communication area includes:

[0038] Identify the back contour of the communication device in the communication area and the hand contour of the person in the communication area, and obtain the time period when the hand contour and the back contour have an intersection as the first judgment time period;

[0039] Identify the head contour of a person in the communication area, and obtain the time period during which the back contour is within the head contour as the second judgment time period;

[0040] Based on the intersection of the first judgment period and the second judgment period, the communication focus duration of each communication area is obtained. Based on the ratio of the communication focus duration to the preset working time, the communication focus level of each communication area is obtained.

[0041] Optionally, in one possible implementation of the first aspect, the step of selecting multiple workstation images within a preset time period based on a video frame selection strategy to obtain the assessment retrospective video for each assessment end includes:

[0042] The recognition area corresponding to the corresponding assessment end in the actual assessment image is used as the display area, and the remaining recognition areas are used as the cover area.

[0043] Based on the corresponding assessment terminal, the corresponding judgment area in the recognition template is determined as the reserved area, and the remaining judgment areas are determined as the hidden area;

[0044] The current data of the assessment terminal is determined to be attendance data, and any workstation image with personnel in the reserved area is selected from the workstation images within the first preset time period as the starting image;

[0045] Select any image of a workstation with personnel in the reserved area within the second preset time period as the final image;

[0046] The current data of the assessment terminal is determined to be absence data, and any workstation image in the retention area that does not have personnel within the first preset time period is selected as the starting image;

[0047] Select any workstation image in the reserved area that does not have personnel within the second preset time period as the termination image. The preset time period includes the first preset time period and the second preset time period.

[0048] The assessment retrospective video for each assessment endpoint is obtained based on the actual assessment image within the preset duration, the starting image, and the ending image.

[0049] Optionally, in one possible implementation of the first aspect, obtaining the assessment retrospective video of each assessment end based on the actual assessment image within a preset duration, the starting image, and the ending image includes:

[0050] A preset masking layer is retrieved to mask the masking area in the actual assessment image to obtain a status frame.

[0051] A preset masking layer is retrieved to mask the image corresponding to the hidden area in the recognition template above the starting image to obtain the starting frame;

[0052] A preset masking layer is retrieved to mask the image corresponding to the hidden area in the recognition template above the termination image to obtain the termination frame;

[0053] By statistically analyzing all status frames, start frames, and end frames of each assessment terminal within a preset time period, the assessment retrospective video of each assessment terminal is obtained.

[0054] A second aspect of the present invention provides an intelligent assessment data processing system, comprising:

[0055] The receiving module is used to receive the workstation template diagram of the target enterprise configured by the management terminal. The workstation template diagram includes multiple identification areas, and each identification area has a corresponding assessment terminal.

[0056] The processing module is used to receive the assessment data uploaded by each of the assessment terminals, and process the corresponding recognition area in the workstation template diagram based on the assessment data to obtain the theoretical assessment diagram.

[0057] The recognition module is used to acquire workstation images collected by the acquisition device, recognize the workstation images based on the assessment recognition strategy and workstation template diagram, obtain updated data of the assessment end, and update the corresponding recognition area of ​​the theoretical assessment diagram based on the updated data to obtain the actual assessment diagram.

[0058] The sending module is used to select and process multiple workstation images within a preset time period based on a video frame selection strategy to obtain the assessment retrospective video of each assessment end, and send the actual assessment image and the assessment retrospective video to the management end.

[0059] The beneficial effects of this invention are as follows:

[0060] 1. This invention provides a visual representation of enterprise performance evaluation data through images, making attendance results more realistic and intuitive. Simultaneously, it generates performance review videos from performance data within a preset timeframe, improving the efficiency of subsequent performance data review. On one hand, this invention maps the evaluation endpoint to the identification areas in the workstation template diagram. Based on the received performance data, the corresponding identification areas are processed to obtain a theoretical evaluation diagram. Then, the collected workstation images are identified using the evaluation recognition strategy and the workstation template diagram to obtain updated data. The corresponding identification areas are then updated to obtain the actual evaluation diagram, making the displayed employee attendance results more intuitive and easier to view. On the other hand, a video frame selection strategy is used to select and process multiple workstation images within a preset time period to obtain performance review videos for each evaluation endpoint. Simultaneously, the actual evaluation diagrams and performance review videos are sent to the management terminal. The performance review videos allow for verification of attendance data within a preset timeframe, reducing employee objections to attendance results and improving the authenticity of attendance data.

[0061] 2. This invention updates the color of the borders of the corresponding recognition areas in the theoretical assessment image based on the assessment data uploaded by employees and the updated data obtained through recognition and judgment, thus obtaining the actual assessment image. This makes the displayed attendance results clearer and more intuitive. First, based on the assessment data uploaded by each assessment terminal, the first pixel value corresponding to the attendance data and the second pixel value corresponding to the absence data are used to update the color of the borders of the corresponding recognition areas in the workstation template image to obtain the theoretical assessment image. Second, based on the range collected by the acquisition device, the corresponding recognition template is located within the workstation template image, and the judgment area in the recognition template is determined. Then, the recognition template is superimposed on the acquired workstation image, and personnel recognition is performed on the image in the judgment area. Specifically, when the number of assessment judgments obtained, i.e., the number of workstation images with personnel in the assessment terminal corresponding to the judgment area, is greater than or equal to a preset number, the current data of the assessment terminal is obtained as attendance data; otherwise, absence data is obtained. Finally, when it is determined that the current data on the assessment end is inconsistent with the assessment data, the color of the border of the recognition area in the corresponding theoretical assessment image is updated using the third pixel value to obtain the actual assessment image, thereby realizing the calibration function of the assessment data and making the assessment data more intuitive. The actual assessment image is obtained by verifying the workstation image, which improves the authenticity and intuitiveness of the attendance data.

[0062] 3. This invention uses a video frame selection strategy to select and process images from multiple workstations to obtain a review video of assessment data for any date from each assessment endpoint. This avoids disputes arising from employee objections to attendance data and improves the efficiency of attendance data review. First, a preset masking layer is used to retain the display area of ​​the actual assessment image corresponding to the assessment endpoint. The remaining masking areas are then masked to obtain a status frame. The preset masking layer is then used to mask the image corresponding to the hidden area in the recognition template above the obtained starting image to obtain a starting frame. The same method is used to obtain the ending frame corresponding to the ending image. All status frames, starting frames, and ending frames from each assessment endpoint within a preset duration are statistically combined to generate an assessment review video. The use of the preset masking layer can protect the attendance information of other employees from being leaked. Secondly, if employees disagree with the received performance evaluation data, they can file an appeal through the evaluation platform. The appeal request can retrieve the collected workstation video as the appeal video based on the appeal date, verifying the employee's attendance record. The performance evaluation retrospective video is then updated accordingly. Specifically, based on the appeal date, the first and second edit frames corresponding to the start and end frames in the performance evaluation retrospective video are replaced with the first and second replacement frames extracted from the appeal video. Simultaneously, the third edit frame corresponding to the status frame is highlighted to indicate that the performance evaluation data has been updated. Finally, the updated performance evaluation retrospective video is sent to the management platform. The generated performance evaluation retrospective video allows for timely review of performance evaluation data, efficient handling of inconsistencies in attendance data, and updating of relevant information, thereby improving the authenticity and reliability of attendance data. Attached Figure Description

[0063] Figure 1 is a schematic diagram of an application scenario provided by an embodiment of the present invention;

[0064] Figure 2 is a flowchart of an intelligent assessment data processing method provided by the present invention;

[0065] Figure 3 is a schematic diagram of the structure of an intelligent assessment data processing system provided by the present invention;

[0066] Figure 4 is a schematic diagram of the hardware structure of an electronic device provided by the present invention. Detailed Implementation

[0067] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0068] Figure 1 illustrates an application scenario provided by an embodiment of the present invention. The present invention extracts a workstation template image based on an internal office environment. By mapping the assessment endpoint to the recognition areas in the workstation template image, and processing the corresponding recognition areas based on the received assessment data, a theoretical assessment image is obtained. Then, the workstation images captured by the camera are identified using an assessment recognition strategy and the workstation template image to determine whether there are personnel at the current workstation and update the data. The corresponding recognition areas are then updated to obtain the actual assessment image. Therefore, assessment data can be displayed in image format. Simultaneously, the captured workstation images are selected and processed according to a video frame selection strategy to obtain assessment retrospective videos for each assessment endpoint. The obtained actual assessment images and assessment retrospective videos are sent to the management endpoint to achieve the storage of accurate attendance information and improve the authenticity of attendance data. The number of office employees is not limited.

[0069] As shown in Figure 2, a first aspect of the present invention provides an intelligent assessment data processing method, comprising S1-S4:

[0070] S1, Receive the workstation template diagram of the target enterprise configured by the management terminal. The workstation template diagram includes multiple recognition areas, and each recognition area has a corresponding assessment terminal.

[0071] In practical applications, displaying clock-in time via spreadsheets makes the assessment data unintuitive. Therefore, it is necessary to first receive the target company's workstation template image configured by the management terminal. The received workstation template image contains multiple recognition areas, and each recognition area also has a corresponding assessment terminal.

[0072] The workstation template is a template of the work locations of company employees. For example, it is a template obtained by dividing the company's office area into employee workstations. The identification area is the area for identifying personnel. The identification area can be the corresponding position area of ​​the employee's workstation in the workstation template. For example, multiple different workstation positions can be assigned corresponding identification areas. Employee A's workstation corresponds to identification area 1 in the workstation template, employee B's workstation corresponds to identification area 2 in the workstation template, and employee C's workstation corresponds to identification area 3 in the workstation template. The assessment terminal can be the employee's corresponding mobile device. The assessment terminal corresponds one-to-one with the identification area. For example, it can be a mobile phone, tablet, etc. Among them, assessment terminal 1 corresponds to identification area 1, assessment terminal 2 corresponds to identification area 2, and assessment terminal 3 corresponds to identification area 3.

[0073] This invention establishes a one-to-one correspondence between the recognition area of ​​the workstation template image and the assessment terminal, making it convenient for personnel to update the corresponding recognition area when uploading assessment data on the assessment terminal. At the same time, the obtained workstation template image also serves as the basis for subsequent workstation image recognition and judgment.

[0074] S2, receive the assessment data uploaded by each of the assessment terminals, and process the corresponding identification area in the workstation template diagram based on the assessment data to obtain the theoretical assessment diagram.

[0075] In practical applications, the corresponding recognition areas in the workstation template diagram need to be processed based on the assessment data uploaded by employees on the assessment terminal to obtain the theoretical assessment diagram.

[0076] Among them, the assessment data is the employee attendance assessment data, and the theoretical assessment image is the assessment image obtained after the assessment data is processed in the recognition area. For example, the recognition area corresponding to the attendance data in the workstation template image is marked in green, and the recognition area corresponding to the absence data is marked in red. The image obtained after marking is the theoretical assessment image.

[0077] In some embodiments, step S2 (receiving the assessment data uploaded by each of the assessment terminals, and processing the corresponding identification area in the workstation template diagram based on the assessment data to obtain the theoretical assessment diagram) includes S21-S22:

[0078] S21, receive assessment data uploaded by each of the assessment terminals, the assessment data including attendance data and absence data, determine a first pixel value based on the attendance data, and determine a second pixel value based on the absence data.

[0079] It is understandable that the attendance and absence data uploaded through the assessment terminal can be used to obtain the assessment data for the corresponding recognition area. In order to distinguish the recognition areas corresponding to attendance and absence, the first pixel value corresponding to attendance data and the second pixel value corresponding to absence data are determined.

[0080] The first pixel value corresponds to the attendance data, and the second pixel value corresponds to the absence data. For example, the first pixel value corresponding to the attendance data can be green, and the second pixel value corresponding to the absence data can be red. There are no restrictions here.

[0081] This invention distinguishes between attendance data and absence data by using a first pixel value and a second pixel value, making the attendance data displayed in the image more intuitive and clear.

[0082] S22, based on the assessment end, update the border of the corresponding recognition area in the workstation template image by using the first pixel value or the second pixel value to obtain the theoretical assessment image.

[0083] It is understandable that upon receiving the assessment data uploaded by the assessment terminal, the border of the recognition area in the workstation template image corresponding to the attendance data is updated by the first pixel using the recognition area corresponding to the assessment terminal and the assessment data, and the border of the recognition area in the workstation template image corresponding to the absence data is updated by the second pixel.

[0084] For example, employee A uploads attendance data on phone number 1, employee B uploads absence data on phone number 2, and employee C uploads attendance data on phone number 3. Employee A's workstation corresponds to recognition area 1 in the workstation template image, employee B's workstation corresponds to recognition area 2 in the workstation template image, and employee C's workstation corresponds to recognition area 3 in the workstation template image. The first pixel value corresponding to the attendance data can be green, and the second pixel value corresponding to the absence data can be red. Therefore, updating the border of recognition area 1 to green, updating the border of recognition area 2 to red, and updating the border of recognition area 3 to green results in an updated theoretical assessment image.

[0085] S3, acquire the workstation image collected by the acquisition device, identify the workstation image based on the assessment recognition strategy and the workstation template image, obtain the updated data of the assessment end, and update the corresponding recognition area of ​​the theoretical assessment image based on the updated data to obtain the actual assessment image.

[0086] In practical applications, because there are instances where personnel leave immediately after clocking in, in order to calibrate the assessment data, workstation images can be collected using a data acquisition device. Based on the assessment recognition strategy and workstation template, the collected workstation images are recognized, and updated data is obtained after recognition. This updated data is then sent to the assessment terminal. Simultaneously, the corresponding recognition areas in the theoretical assessment image are updated based on the updated data to obtain the actual assessment image.

[0087] It is understandable that the acquisition area of ​​the acquisition device is fixed, so the target area can contain multiple acquisition devices. Among them, the acquisition device can be a camera, the update data is the data that updates the recognition area, and the actual assessment image is the assessment image obtained after the recognition area has been updated by the update data.

[0088] In some embodiments, step S3 (acquiring the workstation image collected by the acquisition device, recognizing the workstation image based on the assessment recognition strategy and the workstation template image, obtaining the updated data of the assessment end, and updating the corresponding recognition area in the theoretical assessment image based on the updated data to obtain the actual assessment image) includes S31-S32:

[0089] S31, retrieve the identification template corresponding to the acquisition device in the workstation template image, use the identification area in the identification template as the judgment area, and obtain the workstation image collected by the acquisition device.

[0090] Understandably, to facilitate the recognition and processing of the acquired workstation images, the recognition template corresponding to the acquisition device is located in the workstation template image. This identifies the corresponding area in the recognition template that corresponds to the recognition area in the workstation template image. This area in the recognition template is then used as the judgment area, with a one-to-one correspondence between the recognition area and the judgment area. Simultaneously, the workstation image is acquired through the acquisition device. For example, if the acquisition area of ​​the acquisition device is fixed, the workstation template image can be cropped using this fixed acquisition area to obtain the recognition template. In this case, judgment area 1 in the recognition template corresponds to recognition area 1 in the workstation template image, and judgment area 2 in the recognition template corresponds to recognition area 2 in the workstation template image. Simultaneously, the workstation image is captured by a camera.

[0091] Specifically, the identification template corresponding to the acquisition device is located within the workstation template diagram based on the range of the acquisition device, and the workstation image is the image acquired by the acquisition device.

[0092] S32, the recognition template is superimposed on the workstation image, personnel recognition is performed on the image of the judgment area in the recognition template to obtain updated data, and the corresponding recognition area in the theoretical assessment image is updated based on the updated data to obtain the actual assessment image.

[0093] Understandably, recognizing the collected workstation images requires overlaying a recognition template onto the workstation images. By recognizing personnel in the judgment area of ​​the recognition template, updated data is obtained. Based on the updated data, the corresponding recognition area in the theoretical assessment image is updated to obtain the actual assessment image.

[0094] For example, the employee corresponding to recognition area 1 uploads attendance data on the assessment platform, the employee corresponding to recognition area 2 uploads absence data, and the employee corresponding to recognition area 3 uploads attendance data. By identifying the attendance of the employee corresponding to recognition area 1, the employee corresponding to recognition area 2, and the employee corresponding to recognition area 3, and identifying their absence from the workstation image, and since the uploaded assessment data for recognition areas 2 and 3 is inconsistent with the identified results, updated data for recognition areas 2 and 3 is obtained. The borders of recognition areas 2 and 3 are then updated with different colors to obtain the actual assessment image.

[0095] In some embodiments, step S32 (performing personnel identification on the image of the judgment area in the recognition template to obtain updated data, and updating the corresponding recognition area in the theoretical assessment image based on the updated data to obtain the actual assessment image) includes S321-S325:

[0096] S321, the number of workstation images of personnel corresponding to each judgment area in the recognition template is identified as the assessment judgment quantity.

[0097] In practical applications, workstation images are collected multiple times during working hours, and personnel are identified in the corresponding judgment areas in order to obtain more accurate assessment data. Personnel identification in images can be achieved through existing technologies, such as static headcount statistics, which are existing technologies and will not be elaborated here.

[0098] The number of workstation images containing personnel is used as the assessment judgment quantity. For example, when acquiring workstation images, each time at an hourly interval, 8 workstation images are obtained in one day. Among these 8 workstation images, 8 images corresponding to judgment area 1 contain personnel images, and 6 images corresponding to judgment area 2 contain personnel images. Therefore, the assessment judgment quantity corresponding to judgment area 1 is 8, and the assessment judgment quantity corresponding to judgment area 2 is 6. This is the number of images collected to determine whether a workstation contains an employee. If the number exceeds the preset quantity, it indicates that the employee is in attendance and working. If the employee is not at their seat for a long time, the corresponding number will be lower.

[0099] S322, if the number of assessment judgments is greater than or equal to the preset number, then the current data of the corresponding assessment terminal is determined to be attendance data.

[0100] It is understandable that employees may leave their workstations briefly for reasons such as attending meetings or getting water. As a result, some workstation images may not be able to identify the employees. Therefore, the number of workstation images that can identify the employees is preset to obtain a preset number.

[0101] Specifically, when the number of assessment judgments is greater than or equal to the preset number, the current data of the corresponding assessment terminal is determined to be attendance data. For example, if the preset number is 5, the assessment judgment number corresponding to judgment area 1 is 8, and the assessment judgment number corresponding to judgment area 2 is 6. Since 8>5 and 6>5, the current data of assessment terminal 1 is attendance data, and the current data of assessment terminal 2 is also attendance data.

[0102] S323, if it is determined that the number of assessment judgments is less than the preset number, then the current data of the corresponding assessment terminal is determined to be absence data.

[0103] Understandably, when the number of assessment judgments is less than the preset number, the current data for the corresponding assessment terminal is determined to be absence data. For example, if the preset number is 5, and the assessment judgment number for assessment area 3 is identified as 4, since 4 < 5, the current data for assessment area 3 is absence data.

[0104] S324, if the current data of the assessment terminal is inconsistent with the assessment data, retrieve the third pixel value as the update data, and use the corresponding recognition area of ​​the assessment terminal in the theoretical assessment diagram as the update area.

[0105] Understandably, when the current data obtained from the assessment end through recognition is inconsistent with the uploaded assessment data, the third pixel value can be retrieved as the updated data, and the recognition area corresponding to the assessment end in the theoretical assessment image can be used as the updated area.

[0106] The third pixel value distinguishes it from the first and second pixel values ​​and can be yellow, purple, etc. For example: Employee B uploads absence data to the corresponding recognition area 2 on assessment terminal 2, while Employee C uploads attendance data to the corresponding recognition area 3 on assessment terminal 3. Due to a change in circumstances, Employee B is present that day. Therefore, the personnel identification in the workstation image determines that the current data on assessment terminal 2 is attendance data. However, because Employee C is not present in the corresponding recognition area 3 of the extracted workstation image, the personnel identification in the workstation image determines that the current data on assessment terminal 3 is absence data. Since the assessment data for terminals 2 and 3 is inconsistent with the current data, updated data for recognition areas 2 and 3 is determined. The third pixel value of the border of recognition area 2 is yellow, and the third pixel value of the border of recognition area 3 is purple. Simultaneously, recognition areas 2 and 3 in the theoretical assessment image are the updated areas.

[0107] S325, based on the updated data, update the border of the updated area to obtain the actual assessment image.

[0108] Understandably, the third pixel value is used to update the border of the updated area to obtain the actual assessment image. For example, if the third pixel value of the border of recognition area 2 is determined to be yellow and the third pixel value of the border of recognition area 3 is determined to be purple, the red corresponding to the border of recognition area 2 in the obtained theoretical assessment image is updated to yellow, and the green corresponding to the border of recognition area 3 is updated to purple, thus obtaining the actual assessment image.

[0109] In some embodiments, it also includes:

[0110] A1, retrieve the job attributes from the judgment area of ​​the recognition template, the job attributes include R&D attributes and communication attributes.

[0111] It should be noted that since the job attributes of the personnel corresponding to different workstations are different, for example, some R&D personnel need to use a keyboard for long periods of time to work, while others may need to use tools such as mobile phones for work communication. Therefore, the work situation in different judgment areas can be identified and judged according to different job attributes in order to obtain the work focus of the personnel.

[0112] Among them, the job attribute is the job function attribute of the person corresponding to the judgment area, including the R&D attribute and the communication attribute. The R&D attribute is the function attribute corresponding to the R&D personnel, such as the person corresponding to the judgment area being an R&D personnel. The communication attribute is the attribute of the person corresponding to the judgment area who needs to communicate at work, such as the function attribute corresponding to human resources, customer service, sales, etc.

[0113] Through the above implementation method, the job attributes corresponding to each judgment area can be determined, so as to identify and judge the work focus of personnel in different areas based on the job attributes.

[0114] A2, based on the R&D attributes, the corresponding judgment area is designated as the R&D area, and based on the communication attributes, the corresponding judgment area is designated as the communication area.

[0115] It is understandable that the R&D area is the judgment area corresponding to the R&D attribute, and the communication area is the judgment area corresponding to the communication attribute.

[0116] A3. Perform a continuous focus assessment on the R&D areas to obtain the R&D focus level of each R&D area.

[0117] It is understandable that, given the nature of R&D work, which requires personnel to focus for extended periods, such as long periods of code input or modification, continuous identification and judgment of the R&D area can be performed to obtain the level of R&D focus.

[0118] Among them, R&D focus refers to the level of focus of the personnel in the R&D area.

[0119] In some embodiments, step A3 (the step of performing continuous focus determination on the R&D areas to obtain the R&D focus of each R&D area) includes:

[0120] A31, retrieve the orientation of the display devices in each R&D area, identify the facial orientation of the personnel in the R&D area, and record the time period when the device orientation is opposite to the facial orientation as the first conditional time period for each R&D area.

[0121] It is understandable that when people are focused on their work, they need to face the display device at their corresponding workstation. Therefore, the first conditional time period can be determined based on the orientation of the device and the orientation of their face.

[0122] Among them, the display device refers to the device in the R&D area that displays work content, such as a computer; the device orientation refers to the orientation of the display device's screen, such as the orientation towards the person; the face orientation refers to the orientation of the person's face in the R&D area; and the first conditional time period refers to the time period during which the person is facing the display device.

[0123] It is easy to understand that when the initial condition for judging whether a person is working is that the person needs to face the computer, then when it is recognized that the orientation of the device is opposite to the orientation of the face, it can be said that the person is facing the computer and can be preliminarily indicated that the person is working. Therefore, the time period when the orientation of the device is opposite to the orientation of the face can be used as the first condition period for each R&D area.

[0124] A32, the time period during which personnel in each R&D area trigger input devices is used as the second conditional time period for each R&D area.

[0125] It should be noted that when personnel are focused on their work, they will continuously trigger input devices in the R&D area, such as long-term triggering of the keyboard or mouse, and the corresponding time period can be used as the second condition period.

[0126] The input device is the device used to input work content, such as a keyboard or mouse. The second condition period is the time period during which the person triggers the input device.

[0127] Through the above implementation methods, the present invention can obtain a second conditional time period in order to subsequently determine the corresponding R&D focus.

[0128] A33. Based on the intersection of the first conditional time period and the second conditional time period of each R&D area, the conditional superposition time period of each R&D area is obtained.

[0129] It is understandable that the overlapping period of conditions is the intersection of the first condition period and the second condition period.

[0130] A34. Select an image at any time during the time period corresponding to the conditions overlay for each R&D area as the reference image, and use the images at two times adjacent to the reference image as comparison images.

[0131] It is understandable that when a person's concentration is higher, the range of motion changes within a certain period of time is smaller. Therefore, images can be selected within the overlapping time period to compare the position of the motion, thus obtaining the corresponding R&D concentration.

[0132] The reference image is the image used to compare the magnitude of action changes, i.e., the image within the condition superposition period. The comparison image is the image used to compare actions with the reference image, i.e., the image at two adjacent moments to the reference image.

[0133] A35, select the pixels at the same position in the reference image and the comparison image as the origin of coordinates, and perform coordinate processing on the reference image and the comparison image based on the origin of coordinates.

[0134] Understandably, in order to compare the positions of the baseline image and the comparison image, pixels at the same position are selected as the origin of the coordinates. Generally, the baseline image and the comparison image are processed into coordinates to facilitate subsequent position comparison between the baseline image and the comparison image, so as to obtain the research and development focus.

[0135] A36, determine the pixels with different pixel values ​​in the reference image and the comparison image under the same coordinates as the changed pixels, obtain the number of changes of the changed pixels, and the total number of pixels in the reference image.

[0136] It is understandable that when the pixel values ​​at the same location in two images are inconsistent, it means that the corresponding person's position has changed at adjacent times. Therefore, the pixels at the corresponding locations can be used as changed pixels to count the number of changed pixels and obtain the number of changes, which will help determine the focus of research and development in the future.

[0137] Among them, changing a pixel refers to a pixel with a different pixel value at the same position, changing a number of pixels refers to the number of pixels that are changed, and the total number refers to the total number of corresponding pixels in the base image.

[0138] A37. Based on the ratio of the number of changes to the total number, the change percentage is obtained. When the change percentage is less than the preset change percentage, the moment of the corresponding benchmark image is taken as the R&D focus moment.

[0139] It is understandable that the change percentage is the ratio of the number of changes to the total number, and the preset change percentage is the pre-set percentage of position changes.

[0140] It is easy to understand that when the calculated proportion is less than the preset change proportion, it means that the number of pixels offset between the selected comparison image and the reference image is small, that is, the range of personnel movement is small, which indicates that the personnel are focused on their work. Therefore, the time corresponding to the reference image can be taken as the time of R&D focus.

[0141] Among them, the R&D focus moment refers to the moment when R&D personnel are focused on their work.

[0142] A38, the R&D focus time is obtained by statistically analyzing the R&D focus time, and the R&D focus level of each R&D area is obtained based on the ratio of the R&D focus time to the preset working time.

[0143] Understandably, R&D focus time refers to the amount of time R&D personnel focus on their work, while preset work time refers to the pre-set work time, such as 8 hours.

[0144] It is easy to understand that by selecting the baseline image sequentially during the condition superposition period and selecting adjacent images as comparison images, the R&D focus time is determined, and the R&D focus time is statistically analyzed to obtain the R&D focus duration.

[0145] A4. Intermittent focus assessment is performed on the communication areas to obtain the communication focus level of each communication area.

[0146] Understandably, since the nature of the work corresponding to the communication attribute requires people to make phone calls at irregular intervals, the communication zone can be identified and judged intermittently in order to obtain the communication focus level.

[0147] Among them, communication focus refers to the level of focus of the personnel in the communication area.

[0148] In some embodiments, step A4 (intermittently judging the focus of the communication areas to obtain the communication focus of each communication area) includes:

[0149] A41, identify the back contour of the communication device in the communication area and the hand contour of the person in the communication area, and obtain the time period when the hand contour and the back contour have an intersection as the first judgment time period.

[0150] Understandably, in order to identify whether people in the communication area are engaged in normal work communication, it is possible to identify the intersection of the outline of the person holding the phone and the outline of the person answering the call. When people are communicating normally, they will hold the phone. Therefore, the time period when the outline of the hand and the outline of the back of the person intersect can be used as the first judgment period.

[0151] Among them, the communication device is the device used for work communication, such as a telephone; the back outline is the outline of the device on the back corresponding to the communication device; the hand outline is the outline of the hand of the person in the communication area; and the first judgment period is the time period in which the hand outline and the back outline intersect.

[0152] A42, identify the head contour of the person in the communication area, and obtain the time period when the back contour is within the head contour as the second judgment time period.

[0153] It is understandable that the head outline is the outline of the person's head in the communication area, and the second judgment period is the period when the face outline is within the head outline.

[0154] A43. Based on the intersection of the first judgment period and the second judgment period, the communication focus duration of each communication area is obtained. Based on the ratio of the communication focus duration to the preset working time, the communication focus level of each communication area is obtained.

[0155] It is understandable that the longer the back outline is within the head outline, the longer the person spends communicating at work. Therefore, based on the intersection of the first and second judgment periods, the communication focus time corresponding to the communication area can be obtained, and a ratio can be calculated with the preset work time to obtain the communication focus level.

[0156] Among them, communication focus duration refers to the amount of time that personnel in the communication zone focus on communication, and communication focus level refers to the degree of focus that personnel in the communication zone have on their corresponding work.

[0157] S4, based on the video frame selection strategy, select and process multiple workstation images within a preset time period to obtain the assessment retrospective video of each assessment end, and send the actual assessment image and the assessment retrospective video to the management end.

[0158] In practical applications, to facilitate personnel's review of assessment data, relevant assessment data is collected, retained, and sent to the management terminal.

[0159] Therefore, the obtained workstation images are selected and processed according to the video frame selection strategy to obtain the assessment retrospective video for each assessment end. The obtained actual assessment images and assessment retrospective videos are then sent to the management end. The preset time period is a pre-set time frame for selecting workstation images, such as 8:00-17:00 during the day. This means that the workstation images being selected are those within the preset time period.

[0160] Among them, the assessment review video is a video that allows for review of assessment data. For example, multiple selected assessment data images can be combined into a video.

[0161] In some embodiments, step S4 (selecting multiple workstation images within a preset time period based on a video frame selection strategy to obtain the assessment retrospective video for each assessment end) includes S41-S47:

[0162] S41, obtain the recognition area corresponding to the corresponding assessment end in the actual assessment diagram as the display area, and use the remaining recognition areas as the cover area.

[0163] In practical applications, the data collection device may capture the private information of other employees. Therefore, the following masking scheme was developed to distinguish the attendance data of each assessment terminal and obtain separate assessment review videos for each assessment terminal.

[0164] Understandably, when acquiring attendance data from each assessment terminal, the recognition area corresponding to the respective assessment terminal in the actual assessment image is used as the display area, while the remaining recognition areas are used as masking areas. For example, when selecting and processing the assessment data corresponding to assessment terminal 1 to obtain the assessment review video for assessment terminal 1, recognition area 1 in the actual assessment image corresponding to assessment terminal 1 is used as the display area, while recognition areas 2, 3, and the rest are used as masking areas.

[0165] The present invention further distinguishes the assessment data of each assessment end by using a preset overlay layer, thereby protecting the attendance information of each company employee.

[0166] S42, based on the corresponding assessment terminal, determine the corresponding judgment area in the recognition template as the reserved area, and the remaining judgment areas as the hidden area.

[0167] Understandably, when acquiring attendance data from each assessment terminal, the judgment area corresponding to that assessment terminal in the recognition template is kept as the reserved area, while the remaining judgment areas are hidden areas. For example, judgment area 1 in the recognition template corresponding to assessment terminal 1 is kept as the reserved area, while judgment areas 2, 3, and the rest are hidden areas.

[0168] S43, determine that the current data of the assessment terminal is attendance data, and select any workstation image with personnel in the reserved area from the workstation images within the first preset time period as the starting image.

[0169] Understandably, when the current data of the assessment end is determined to be attendance data after processing, a workstation image with personnel is selected from the workstation images obtained within the first preset time period, and this workstation image is used as the starting image.

[0170] The first preset time period is a time period that is determined in advance by humans, such as 8:00 AM to 12:00 PM.

[0171] For example: if the No. 1 assessment terminal corresponding to the No. 1 recognition area is determined to be the attendance data, select one image of a person at a workstation in the No. 1 retention area between 8:00 AM and 12:00 PM, and use this selected workstation image as the starting image.

[0172] S44, select any image of a workstation with personnel in the reserved area within the second preset time period as the termination image.

[0173] Understandably, when the current data of the assessment end is determined to be attendance data after processing, a workstation image with personnel is selected from the workstation images obtained within the second preset time period, and this workstation image is used as the termination image.

[0174] The second preset time period is a time period that is determined in advance by humans, such as 14:00-17:00.

[0175] For example: if the No. 1 assessment terminal corresponding to the No. 1 recognition area is determined to be the attendance data, select an image of a workstation with someone at the seat in the No. 1 retention area between 14:00 and 17:00, and select this workstation image as the termination image.

[0176] S45, determine that the current data of the assessment terminal is absence data, and select any workstation image in the retention area that does not have personnel within the first preset time period as the starting image.

[0177] Understandably, when the current data of the assessment end is determined to be absent data after processing, a workstation image without personnel is selected from the workstation images obtained within the first preset time period, and this workstation image is used as the starting image.

[0178] For example: if the No. 3 assessment terminal corresponding to the No. 3 recognition area is determined to be absent data, select an image of a workstation with no one in the seat from the No. 3 retention area between 8:00 am and 12:00 pm, and use this selected workstation image as the starting image.

[0179] S46, select any workstation image in the reserved area that does not have personnel within the second preset time period as the termination image, the preset time period includes the first preset time period and the second preset time period.

[0180] Understandably, when the current data of the assessment end is determined to be absent data after processing, a workstation image without personnel is selected from the workstation images obtained within the second preset time period, and this workstation image is used as the termination image.

[0181] For example: if the No. 3 assessment terminal corresponding to the No. 3 recognition area is determined to be absent data, select an image of a workstation with no one in the seat from 14:00 to 17:00 in the No. 3 retention area, and select this workstation image as the termination image.

[0182] S47, based on the actual assessment chart within the preset time period, the starting chart, and the ending chart, obtain the assessment retrospective video for each assessment end.

[0183] Understandably, the assessment review video consists of the actual assessment chart, the starting chart, and the ending chart within a preset duration.

[0184] The preset duration is manually set, for example, one month. The assessment retrospective video for each assessment terminal consists of all actual assessment images, starting images, and ending images within the preset duration. For example, the assessment retrospective video for assessment terminal 1 is a video composed of the actual assessment images, starting images, and ending images for each day of recognition area 1 within a continuous statistical period of one month.

[0185] In some embodiments, step S47 (obtaining the assessment retrospective video of each assessment end based on the actual assessment image within a preset duration, the starting image, and the ending image) includes S471-S474:

[0186] S471, retrieve the preset masking layer and mask the masking area in the actual assessment image to obtain a status frame.

[0187] Understandably, to protect the assessment data of company employees from being leaked in the actual assessment chart, only the display area information in the actual assessment chart is shown, that is, only the assessment data of the corresponding assessment end is displayed, and the masking area is masked. Therefore, it is necessary to retrieve the preset masking layer. The preset masking layer is a layer that is preset by the user for masking, such as a mosaic layer. The masking area in the actual assessment chart is masked to obtain the status frame. For example, if the first recognition area in the actual assessment chart is the display area, the masking area in the actual assessment chart is masked using the preset masking layer, and the resulting masked actual assessment chart is used as the status frame.

[0188] S472, retrieve the preset masking layer and mask the image corresponding to the hidden area in the recognition template above the starting image to obtain the starting frame.

[0189] Understandably, the preset masking layer is retrieved to mask the recognition template above the starting image, the image corresponding to the reserved area in the recognition template in the starting image is retained, and the image corresponding to the hidden area in the recognition template above the starting image is masked. The resulting masked starting image is the starting frame.

[0190] For example: Obtain the starting image corresponding to workstation 1, place the recognition template on this starting image, and obtain the 1st judgment area as the reserved area, and the other judgment areas 2, 3, etc. as the hidden areas. Use the preset masking layer to mask the corresponding images in the masking area through the recognition template, and obtain the masked starting image as the starting frame.

[0191] S473, retrieve the preset masking layer and mask the image corresponding to the hidden area in the recognition template above the termination image to obtain the termination frame.

[0192] Understandably, the preset masking layer is retrieved to mask the recognition template above the termination image, the image corresponding to the reserved area in the recognition template in the termination image is retained, and the image corresponding to the hidden area in the recognition template above the termination image is masked. The resulting masked termination image is the termination frame.

[0193] For example: If the termination image corresponding to workstation 1 is obtained, the recognition template is placed on this termination image, and the judgment area 1 is the reserved area, while the other judgment areas 2, 3, etc. are the hidden areas. The preset masking layer is used to mask the corresponding images in the masking area through the recognition template, and the masked termination image is the termination frame.

[0194] S474, count all status frames, start frames and end frames of each assessment terminal within the preset time period to obtain the assessment backtracking video of each assessment terminal.

[0195] It is understandable that the assessment review video consists of all the status frames, start frames, and end frames of each assessment terminal within a preset time period.

[0196] The assessment retrospective video for each assessment endpoint is a video composed of the actual assessment image, start image, and end image within a preset duration, after being masked to obtain the corresponding status frame, start frame, and end frame. For example, starting from October 1st, the status frame, start frame, and end frame of each assessment endpoint are counted daily, and the daily status frame, start frame, and end frame are combined to form the daily assessment retrospective video. The status frame, start frame, and end frame of each assessment endpoint are counted continuously for a month, and the daily assessment retrospective videos are combined to form the monthly assessment retrospective video.

[0197] Based on the above embodiments, the present invention also includes B1-B2:

[0198] B1, bind the assessment data uploaded by the assessment terminal with the corresponding recognition area in the actual assessment diagram.

[0199] It is understandable that the recognition areas on the assessment terminal and the actual assessment diagram are in one-to-one correspondence, so the assessment data uploaded by the assessment terminal can be bound one-to-one with the corresponding recognition areas on the actual assessment diagram.

[0200] B2, determine that the management terminal triggers any recognition area in the actual assessment diagram, retrieves the assessment data corresponding to the corresponding recognition area and displays it.

[0201] Understandably, the management side can view the assessment data corresponding to any recognition area. Therefore, when the management side triggers any recognition area in the actual assessment chart, the assessment data corresponding to that area will be retrieved and displayed. For example, if employee number 1 uploads attendance data, and the current data is determined after processing, the border of recognition area number 1 in the actual assessment chart is green. If the manager needs to view the assessment data corresponding to employee number 1, the manager can click on recognition area number 1 in the actual assessment chart on the management side to trigger the display of the attendance data for recognition area number 1.

[0202] Based on the above embodiments, the present invention also includes C1-C3:

[0203] C1 receives the appeal request from the assessment end, parses the appeal request to obtain the appeal date, and designates the corresponding assessment end as the appeal end and the identification area corresponding to the appeal end as the appeal area.

[0204] In real-world scenarios, unforeseen circumstances may occur, such as employees attending meetings or fetching water. This can lead to discrepancies in the data obtained when identifying and judging workstation images extracted within a preset time period. To protect the accuracy of employee performance data, employees are given the right to appeal through the performance evaluation system. For example, if eight workstation images are taken every hour within a day, but the employee is absent due to a meeting or fetching water, the number of workstation images for that employee might be only four, less than the preset five. Consequently, the actual performance evaluation image might classify the corresponding recognition area as absent. Therefore, if an employee disagrees with their performance data, they can submit an appeal request through the performance evaluation system.

[0205] Understandably, the appeal request can be parsed to determine the appeal date, which corresponds to the date when the performance evaluation data became inconsistent. The assessment server that submitted the appeal request is designated as the appeal server, and the corresponding recognition area of ​​that server is designated as the appeal area. For example, if assessment server #3 receives performance evaluation data and finds that the border of recognition area #3 in the actual performance evaluation chart for October 4th is purple, indicating that attendance data was uploaded, it is determined to be absent data. Employee C objects to this, therefore, employee C submits an appeal request. Thus, assessment server #3 becomes the appeal server, and recognition area #3 in the actual performance evaluation chart becomes the appeal area.

[0206] C2, based on the appeal area, determine the corresponding acquisition device as the appeal device, and obtain the workstation video acquired by the appeal device as the appeal video according to the appeal date.

[0207] Understandably, the appeal area has corresponding data collection devices. These devices will be designated as the appeal devices, and the workstation video captured by these devices will be used as the appeal video based on the appeal date. For example, if camera A captured workstation 3, the video captured by camera A at workstation 10.4 will be used as the appeal video.

[0208] C3. Update the assessment retrospective video of the appeal terminal based on the appeal video, and send the updated assessment retrospective video to the management terminal.

[0209] Understandably, the appeal video is used to update the corresponding performance review video on the appeal side, and the updated performance review video is then sent to the management side.

[0210] In some embodiments, step C3 (updating the assessment retrospective video of the appealing party based on the appeal video, and sending the updated assessment retrospective video to the management terminal) includes C31-C35:

[0211] C31, based on the date of the appeal, locate the corresponding start frame in the assessment review video of the appealing party as the first edit frame, the corresponding end frame as the second edit frame, and the corresponding status frame as the third edit frame.

[0212] Understandably, the appeal date can be used to locate the start frame, end frame, and status frame in the corresponding assessment review video. The start frame can then be used as the first edit frame, the end frame as the second edit frame, and the status frame as the third edit frame. For example, using the appeal date of October 4th, the start frame of October 4th in the assessment review video corresponding to appeal #3 can be used as the first edit frame, the end frame as the second edit frame, and the status frame as the third edit frame.

[0213] C32, highlight the display area in the third edit frame.

[0214] Understandably, the display area in the third editing frame is highlighted to facilitate subsequent viewing by the administrator.

[0215] C33, based on the first preset time period, extract the workstation image of the person in the appeal area of ​​the appeal video as the first replacement frame.

[0216] Understandably, within the first preset time period in the appeal video, images of workstations with personnel in the appeal area are extracted as the first replacement frame.

[0217] The first replacement frame is the image that replaces the first edit frame. For example, in the obtained video of workstation 10.4, an image of workstation 3 with personnel is extracted between 8:00 AM and 12:00 PM. This workstation image is the first replacement frame.

[0218] C34, based on the second preset time period, extract the workstation image of the person in the appeal area of ​​the appeal video as the second replacement frame.

[0219] It is understandable that images of workstations with personnel in the appeal area are extracted as the second replacement frames within the second preset time period in the appeal video.

[0220] The second replacement frame is the image that replaces the second edit frame. For example, in the obtained video of workstation 10.4, the image of workstation 3 with personnel is extracted between 14:00 and 17:00, and this workstation image is the second replacement frame.

[0221] C35, the first edit frame is replaced and updated according to the first replacement frame, and the second edit frame is replaced and updated according to the second replacement frame, and the updated assessment retrospective video is sent to the management terminal.

[0222] Understandably, the process involves replacing the first edit frame in the assessment retrospective video with the first replacement frame, and the second edit frame with the second replacement frame, to obtain an updated assessment retrospective video, which is then sent to the management terminal. For example, replacing images without personnel in the assessment retrospective video corresponding to assessment terminal 3 (10.4) with images containing personnel, and then sending the updated assessment retrospective video to the management terminal for storage.

[0223] To better implement the intelligent assessment data processing method provided by this invention, this invention also provides an intelligent assessment data processing system, as shown in Figure 3, comprising:

[0224] The receiving module is used to receive the workstation template diagram of the target enterprise configured by the management terminal. The workstation template diagram includes multiple recognition areas, and each recognition area has a corresponding assessment terminal.

[0225] The processing module is used to receive the assessment data uploaded by each of the assessment terminals, and process the corresponding recognition area in the workstation template diagram based on the assessment data to obtain the theoretical assessment diagram.

[0226] The recognition module is used to acquire workstation images collected by the acquisition device, recognize the workstation images based on the assessment recognition strategy and workstation template diagram, obtain updated data of the assessment end, and update the corresponding recognition area of ​​the theoretical assessment diagram based on the updated data to obtain the actual assessment diagram.

[0227] The sending module is used to select and process multiple workstation images within a preset time period based on a video frame selection strategy to obtain the assessment retrospective video of each assessment end, and send the actual assessment image and the assessment retrospective video to the management end.

[0228] Figure 4 shows a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. The electronic device 40 includes: a processor 41, a memory 42, and a computer program; wherein...

[0229] The memory 42 is used to store the computer program, and the memory may also be flash memory. The computer program is, for example, an application program or functional module that implements the above method.

[0230] The processor 41 is configured to execute the computer program stored in the memory to implement the various steps performed by the device in the above method. For details, please refer to the relevant descriptions in the preceding method embodiments.

[0231] Alternatively, the memory 42 can be either standalone or integrated with the processor 41.

[0232] When the memory 42 is a device independent of the processor 41, the device may further include:

[0233] Bus 43 is used to connect the memory 42 and the processor 41.

[0234] The present invention also provides a readable storage medium storing a computer program, which, when executed by a processor, is used to implement the methods provided in the various embodiments described above.

[0235] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for intelligent assessment data processing, characterized in that, include: The system receives a workstation template image of the target enterprise configured by the management terminal. The workstation template image includes multiple identification areas, and each identification area has a corresponding assessment terminal. The system receives assessment data uploaded by each assessment terminal, processes the corresponding recognition areas in the workstation template image based on the assessment data to obtain a theoretical assessment image; acquires workstation images collected by a data acquisition device, recognizes the workstation images based on the assessment recognition strategy and the workstation template image to obtain updated data from the assessment terminal, and updates the corresponding recognition areas in the theoretical assessment image based on the updated data to obtain an actual assessment image. This includes: retrieving the recognition template corresponding to the data acquisition device in the workstation template image, using the recognition areas in the recognition template as judgment areas, and acquiring the workstation images collected by the data acquisition device; overlaying the recognition template onto the workstation image, performing personnel recognition on the images in the judgment areas of the recognition template, and obtaining updated data. The actual assessment image is obtained by updating the corresponding recognition area in the theoretical assessment image based on the updated data, including: identifying the number of workstation images with personnel corresponding to each judgment area in the recognition template as the assessment judgment quantity; determining that the assessment judgment quantity is greater than or equal to a preset quantity, then determining the current data of the corresponding assessment end as attendance data; determining that the assessment judgment quantity is less than the preset quantity, then determining the current data of the corresponding assessment end as absence data; if the current data of the assessment end is inconsistent with the assessment data, retrieving the third pixel value as the update data, and using the recognition area of ​​the corresponding assessment end in the theoretical assessment image as the update area; updating the border of the update area based on the updated data to obtain the actual assessment image. The assessment chart also includes: retrieving the job attributes of the judgment areas in the recognition template, the job attributes including R&D attributes and communication attributes; designating the corresponding judgment areas as R&D areas based on the R&D attributes, and designating the corresponding judgment areas as communication areas based on the communication attributes; performing continuous focus judgment on the R&D areas to obtain the R&D focus of each R&D area, including: retrieving the device orientation of the display devices in each R&D area, identifying the facial orientation of the personnel in the R&D area, and recording the time period when the device orientation and the facial orientation are opposite as the first conditional time period for each R&D area; calculating the time period when the personnel in each R&D area trigger the input device as the second conditional time period for each R&D area; and based on the intersection of the first conditional time period and the second conditional time period for each R&D area, The following steps are taken: Obtain the overlay time periods for each R&D area; select an image from any time within the overlay time periods for each R&D area as a reference image, and use the images from two adjacent times as comparison images; select pixels at the same position in the reference and comparison images as the origin, and perform coordinate transformation on the reference and comparison images based on the origin; determine pixels with different pixel values ​​in the reference and comparison images at the same coordinates as changed pixels, obtain the number of changes to the changed pixels, and the total number of pixels in the reference image; obtain the change ratio based on the ratio of the number of changes to the total number, and determine that when the change ratio is less than a preset change ratio, the time of the corresponding reference image is taken as the R&D focus time.The research and development focus time is statistically analyzed to obtain the research and development focus duration. Based on the ratio of the research and development focus duration to the preset working time, the research and development focus level of each research and development area is obtained. Intermittent focus level judgment is performed on the communication area to obtain the communication focus level of each communication area, including: identifying the back contour of communication devices in the communication area and the hand contour of personnel in the communication area, and obtaining the time period when the hand contour and the back contour intersect as the first judgment period; identifying the head contour of personnel in the communication area, and obtaining the time period when the back contour is within the head contour as the second judgment period; based on the intersection of the first judgment period and the second judgment period, the communication focus duration of each communication area is obtained, and based on the ratio of the communication focus duration to the preset working time, the communication focus level of each communication area is obtained; multiple workstation images within the preset time period are selected and processed based on a video frame selection strategy to obtain the assessment retrospective video of each assessment end, and the actual assessment image and the assessment retrospective video are sent to the management end, including: obtaining... The identification area corresponding to the assessment terminal in the actual assessment image is taken as the display area, and the remaining identification areas are taken as the masking area; the judgment area in the identification template corresponding to the assessment terminal is determined as the retention area, and the remaining judgment areas are taken as the hidden area; the current data of the assessment terminal is determined to be attendance data, and any workstation image with personnel in the retention area within the first preset time period is selected as the starting image; any workstation image with personnel in the retention area within the second preset time period is selected as the ending image; the current data of the assessment terminal is determined to be absence data, and any workstation image without personnel in the retention area within the first preset time period is selected as the starting image; any workstation image without personnel in the retention area within the second preset time period is selected as the ending image, where the preset time period includes the first preset time period and the second preset time period; the assessment retrospective video of each assessment terminal is obtained based on the actual assessment image within the preset duration, the starting image, and the ending image.

2. The method according to claim 1, characterized in that, The step of receiving assessment data uploaded by each of the assessment terminals and processing the corresponding recognition areas in the workstation template image based on the assessment data to obtain a theoretical assessment image includes: receiving assessment data uploaded by each of the assessment terminals, the assessment data including attendance data and absence data; determining a first pixel value based on the attendance data and a second pixel value based on the absence data; and updating the borders of the corresponding recognition areas in the workstation template image based on the first pixel value or the second pixel value from the assessment terminals to obtain a theoretical assessment image.

3. The method according to claim 1, characterized in that, The step of obtaining the assessment retrospective video for each assessment end based on the actual assessment image, the starting image, and the ending image within a preset time period includes: retrieving a preset masking layer to mask the masked area in the actual assessment image to obtain a status frame; retrieving a preset masking layer to mask the image corresponding to the hidden area in the recognition template above the starting image to obtain a starting frame; retrieving a preset masking layer to mask the image corresponding to the hidden area in the recognition template above the ending image to obtain a ending frame; and counting all status frames, starting frames, and ending frames of each assessment end within a preset time period to obtain the assessment retrospective video for each assessment end.

4. An intelligent assessment data processing system, characterized in that, include: The receiving module is used to receive the workstation template diagram of the target enterprise configured by the management terminal. The workstation template diagram includes multiple identification areas, and each identification area has a corresponding assessment terminal. The processing module receives assessment data uploaded by each assessment terminal, processes the corresponding recognition areas in the workstation template image based on the assessment data, and obtains a theoretical assessment image. The recognition module acquires workstation images collected by the acquisition device, recognizes the workstation images based on the assessment recognition strategy and the workstation template image, obtains updated data from the assessment terminal, and updates the corresponding recognition areas in the theoretical assessment image based on the updated data to obtain an actual assessment image. This includes: retrieving the recognition template corresponding to the acquisition device in the workstation template image, using the recognition areas in the recognition template as judgment areas, and acquiring the workstation images collected by the acquisition device; overlaying the recognition template onto the workstation image, and performing image processing on the judgment areas in the recognition template. The process involves identifying personnel, obtaining updated data, and updating the corresponding recognition area in the theoretical assessment image based on the updated data to obtain the actual assessment image. This includes: identifying the number of workstation images with personnel at the assessment end corresponding to each judgment area in the recognition template as the assessment judgment quantity; determining that the assessment judgment quantity is greater than or equal to a preset quantity, then determining the current data of the corresponding assessment end as attendance data; determining that the assessment judgment quantity is less than the preset quantity, then determining the current data of the corresponding assessment end as absence data; if the current data of the assessment end is inconsistent with the assessment data, retrieving the third pixel value as updated data, and using the recognition area of ​​the corresponding assessment end in the theoretical assessment image as the updated area; and updating the border of the updated area based on the updated data. The new processing yields the actual assessment chart; it also includes: retrieving the job attributes of the judgment areas in the recognition template, the job attributes including R&D attributes and communication attributes; based on the R&D attributes, designating the corresponding judgment areas as R&D areas, and based on the communication attributes, designating the corresponding judgment areas as communication areas; performing continuous focus judgment on the R&D areas to obtain the R&D focus of each R&D area, including: retrieving the device orientation of the display devices in each R&D area, identifying the facial orientation of the personnel in the R&D area, and recording the time period when the device orientation and the facial orientation are opposite as the first conditional time period for each R&D area; statistically analyzing the time periods when personnel in each R&D area trigger the input devices as the second conditional time period for each R&D area; and based on the first conditional time period and the second conditional time period for each R&D area... The intersection of the conditions is used to obtain the time periods for the conditional overlay of each R&D area; the image of any time in the conditional overlay time period corresponding to each R&D area is selected as the reference image, and the images of the two times adjacent to the reference image are used as comparison images; the pixels at the same position in the reference image and the comparison image are selected as the origin of the coordinate system, and the reference image and the comparison image are processed by coordinate transformation based on the origin of the coordinate system; the pixels with different pixel values ​​in the reference image and the comparison image under the same coordinate system are determined as the changed pixels, and the number of changes of the changed pixels and the total number of pixels in the reference image are obtained; the change ratio is obtained according to the ratio of the number of changes to the total number, and when the change ratio is less than the preset change ratio, the time of the corresponding reference image is taken as the R&D focus time;The following steps are performed: First, the R&D focus time is calculated by statistically analyzing the R&D focus moments. Second, the R&D focus duration is obtained. Third, the R&D focus level of each R&D area is calculated based on the ratio of the R&D focus duration to the preset working time. Fourth, the communication focus level of each communication area is determined by intermittent focus assessment, including: identifying the back contour of communication devices and the hand contour of personnel in the communication area, and using the time period when the hand contour and the back contour intersect as the first assessment time period; identifying the head contour of personnel in the communication area, and using the time period when the back contour is within the head contour as the second assessment time period; using the intersection of the first and second assessment time periods to obtain the communication focus duration of each communication area, and using the ratio of the communication focus duration to the preset working time to obtain the communication focus level of each communication area. Fifth, a sending module is used to select and process multiple workstation images within a preset time period based on a video frame selection strategy to obtain assessment retrospective videos for each assessment end, and send the actual assessment image and the assessment retrospective video to the management end. This includes: obtaining the recognition area corresponding to the assessment terminal in the actual assessment image as the display area, and using the remaining recognition areas as the masking area; determining the judgment area corresponding to the recognition template based on the corresponding assessment terminal as the retention area, and using the remaining judgment areas as the hidden area; determining that the current data of the assessment terminal is attendance data, selecting any workstation image with personnel in the retention area within a first preset time period as the starting image; selecting any workstation image with personnel in the retention area within a second preset time period as the ending image; determining that the current data of the assessment terminal is absence data, selecting any workstation image without personnel in the retention area within the first preset time period as the starting image; selecting any workstation image without personnel in the retention area within the second preset time period as the ending image, wherein the preset time period includes the first preset time period and the second preset time period; and obtaining the assessment retrospective video of each assessment terminal based on the actual assessment image, the starting image, and the ending image within the preset duration.

Images