CN118863826A - An evaluation method and system based on image calibration - Google Patents

Abstract

The present invention discloses an evaluation method and system based on image calibration, and the method includes: S1, drawing an evaluation template, creating an image mark, and obtaining an evaluation template file with an image mark; S2, based on the evaluation template file, obtaining a marked evaluation template image and a standard evaluation template image; S3, calculating the template coordinates based on the marked evaluation template image, and obtaining a first evaluation frame center point coordinate set; S4, based on the first evaluation frame center point coordinate set, taking the standard evaluation template image as the base map, loading a custom evaluation frame control, completing template mapping, and obtaining a template index library; S5, maintaining the evaluation theme, selecting the evaluation template, printing, distributing, and filling in the evaluation form; S6, collecting the filled evaluation form, scanning the evaluation form image, identifying the filled score, and obtaining the evaluation result. By creating an image mark on the evaluation template, and obtaining a template index library through identification and mapping based on image calibration, the evaluation template mapping efficiency and the evaluation form recognition efficiency can be improved.

Description

An evaluation method and system based on image calibration

Technical Field

The present invention relates to the field of assessment, and in particular to an assessment method and system based on image calibration applied to human resources.

Background Art

Investigating, collecting information, and conducting assessments and evaluations are important means and tools for modern enterprises to optimize production and operation activities, enhance product and service performance, optimize organizational structure, improve employee skills, and realize corporate value.

Personnel assessment is an important part of human resource management, involving the use of a variety of techniques and methods to measure and evaluate an individual's abilities, skills and other work-related attributes. The purpose of personnel assessment is to better understand an individual's qualities and potential, so as to make more effective decisions in subsequent human resource activities.

With the development of information technology, the implementation methods of personnel assessment have become more diversified and efficient. The introduction of information technology into the traditional manual assessment method has realized semi-manual assessment. Data collection is achieved through the Internet, and data analysis technology is introduced. It is a method commonly used by personnel departments of organizations at all levels. In actual application scenarios, many companies still use paper evaluation forms to conduct evaluations based on the introduction of information systems. Although this semi-manual evaluation method can reduce the evaluation intensity of the human resources department to a certain extent, there are still many problems in the production of paper evaluation forms, information identification, and data processing. For example, when marking the evaluation template, it is necessary to manually mark each page in the information system to bind the evaluation indicators and personnel; for the same evaluation template, for personnel from different evaluation units, it is necessary to repeat the marking one by one, and if there are omissions or wrong markings, rework and re-marking are required. Although the information system is introduced for auxiliary evaluation, the work of human resources evaluation staff is still cumbersome, complicated, and inefficient; in addition, when identifying the results of filling in the paper, most of the corresponding data is found in the template through the filled-in points, the recognition speed is slow, and the accuracy needs to be improved, which affects the overall efficiency of the evaluation work.

Summary of the invention

To solve the above problems, the present invention provides an evaluation method and system based on image calibration, which aims to calculate template coordinates through image marking, and utilize the automatic loading of evaluation box controls in the evaluation system to complete batch labeling of horizontal and vertical evaluation boxes of the evaluation template, and generate a template indicator library; when filling in the evaluation form, based on the template indicator library, quickly lock the filling position, identify the filling score, and analyze the evaluation result; a template is made once to support multiple repeated uses, which will greatly improve the efficiency of marking evaluation forms and the recognition efficiency of evaluation results, including those used in personnel ability evaluation, and improve the efficiency of evaluation decision-making support.

To achieve the above object, the present invention provides an evaluation method based on image calibration, the method comprising the following steps:

S1. Draw an evaluation template, create an image tag, and obtain an evaluation template file with an image tag;

S2. Based on the evaluation template file, obtain the marked evaluation template image and the standard evaluation template image;

S3, calculating the template coordinates based on the marked evaluation template image to obtain a coordinate set of the center point of the first evaluation frame;

S4. Based on the coordinate set of the center point of the first evaluation frame, the custom evaluation frame control is automatically loaded, the template is plotted, and a template indicator library is obtained;

S5. Maintain the evaluation theme, select the evaluation template, print, distribute and fill in the evaluation form;

S6. Collect the filled evaluation form, scan the evaluation form image, identify the filled score, and obtain the evaluation result.

Preferably, in step S1, the evaluation template is a two-dimensional table including evaluation objects and evaluation indicators, and the two-dimensional table contains a fill-in area for evaluation, and the fill-in area is set as a connected domain block of preset evaluation options.

Preferably, in step S1, the image mark is a connected domain color block configured on the evaluation template, which is placed outside the two-dimensional table and corresponds to the filled area in the evaluation template, and is used as a mark for calculating the coordinates of the filled area. More preferably, the image mark is a connected domain color block configured on the evaluation template, which is placed outside the two-dimensional table and corresponds to the shape and size of the filled area in the evaluation template, including row marks and column marks, which are used for calculating the coordinates of the center point of the filled area.

Preferably, in step S1, the evaluation template file with image marks includes an evaluation title, a QR code, an evaluation unit, a weight, a two-dimensional table, instructions (i.e., notes), four corner positioning points, and an image mark; the four corner positioning points are connected domains located at the four corners of the paper (upper left corner, lower left corner, upper right corner, and lower right corner), and each positioning point has a certain area and a center point; all contents in the evaluation template file are located in a rectangular frame area formed by connecting the center points of the four corner positioning points; the weight refers to the sample weight, which is used to identify and distinguish the category attributes of the sample, so as to adjust the proportion or proportional distribution of the sample in different groups participating in the collection of evaluation information when the total amount of evaluation samples is the same.

Preferably, in step S2, the set of all filled areas in the two-dimensional table in the marked evaluation template image constitutes a first evaluation frame, and the coordinate set of the center points of all filled areas constitutes a coordinate set of the center points of the first evaluation frame; the set of all filled areas in the two-dimensional table in the standard evaluation template image constitutes a second evaluation frame; the marked evaluation template image and the standard evaluation template image are obtained by printing and scanning.

Preferably, in step S3, the template coordinate calculation is to perform image standardization processing on the evaluation template picture with image mark, and then perform identification and calculation to determine the positioning point and the image mark position and correct the image; take the coordinates of the center point of the positioning point in the upper left corner of the image as the origin O(0,0), calculate the coordinates of the four corner points of the two-dimensional table and the coordinates of the center point of the image mark, perform two-dimensional table correction, and use the upper right corner point of the two-dimensional table as a reference to identify the image mark position, take the _Xi value of the center point coordinate _Ai ( _Xi , _Ya ) of the upper image mark, and the _Yi value of the center point coordinate _Ri ( _Xb , _Yi ) of the right image mark, and combine them to obtain the first evaluation frame center point coordinate set S _first ={ _C1 ( _X1 , _Y1 ), _C2 ( _X2 , _Y2 ),…, _Cn ( _Xn , _Yn )}; the image standardization processing is to perform grayscale conversion, binary segmentation and opening operation on the image.

Preferably, in step S4, the template marking is a process of selecting and marking the third evaluation frame using a mouse or a touch device;

The third evaluation frame uses the standard evaluation template image after image standardization and recognition calculation as the base map, takes the upper left corner positioning point of the base map as the origin, and uses the first evaluation frame center point coordinate set S _first to load a set of custom evaluation frame controls that can completely cover the second evaluation frame (the set of all filled areas in the standard evaluation template image) on the base map; the automatic loading of the third evaluation frame does not require the production and identification of the fully filled evaluation template, and the filled areas with various attribute information can be obtained as evaluation frame controls for template marking; the template calculation results can also be verified by the coverage of the third evaluation frame on the second evaluation frame. If the deviation is large, it is necessary to re-scan and recalculate; the custom evaluation frame control attributes include score, score color, score font size, drawing line width, height, drawing line color, belonging indicator, belonging object, unique identifier, and center point coordinates;

The box selection is to select a rectangular area of the third evaluation box by dragging with a mouse or touch device. The third evaluation box is blue by default. When selected, the box selection is displayed on the interface through a black dotted line connected area box to realize the visualization of the box selection content; after the box selection is completed, the evaluation box control of the selected third evaluation box changes from blue to red, and the black dotted line connected area box disappears; it supports arbitrary box selection of a row, a column or a cell of a two-dimensional table;

The annotation realizes batch annotation or single annotation of the framed area data; after completing the horizontal and vertical frame selection and annotation of all the evaluation box controls in the third evaluation box, a template indicator library can be obtained, and the template indicator library attributes include evaluation box control attributes, scores, evaluated objects, and evaluation indicators;

The evaluation frame control attributes include score, score color, score font size, drawing line width, height, drawing line color, belonging indicator, belonging object, unique identifier and center point coordinates.

Preferably, in step S6, the evaluation paper recognition is that after the evaluation form is filled in, it is uniformly collected and connected to a scanner. The evaluation paper is first scanned to obtain an evaluation form image. The computer uses image processing technology to perform image standardization on the evaluation form image, and then searches for the corresponding evaluation template according to the QR code in the current evaluation form image to obtain a template indicator library; OCR recognition and calculation are used to obtain the coordinates of the center points of the four corner positioning points, and rectangular cutting is performed after the perspective transformation of the image is completed. The upper left corner positioning point is used as the origin, and the query range is expanded according to the coordinate parameters in the template indicator library. The coordinates of the starting point and the end point of the query range are calculated, the current coordinate area is cut, and the filled pixel point calculation is performed. It is determined whether it is a filled coordinate based on the ratio. If it is a filled coordinate, the score data of the evaluation indicator and the object corresponding to the current filled coordinate is obtained according to the template indicator library; during the recognition process, for evaluation form images with abnormal filling, such as over-filling, missing filling, and artificial abnormal filling, manual processing is performed item by item, and manual adjustment and saving of scores are supported, as well as discarding and recalculation.

An evaluation system based on image calibration, comprising:

The template drawing module is used to draw the evaluation template and dynamically configure the template content to obtain the evaluation template file with image tags;

An intelligent template marking module is used to calculate the template coordinates to obtain the coordinate set S _first of the center point of the first evaluation frame, and based on S _first , load a custom evaluation frame control as the third evaluation frame on the standard evaluation template image, select and mark the frame, complete the template marking, and generate a template indicator library;

The evaluation module is used to manage evaluation templates, evaluation template indicators, evaluation topics, and evaluated objects, complete the printing of evaluation forms for all evaluated objects, and conduct evaluations; collect, scan, and identify filled evaluation forms, perform calculations based on the template indicator library, and obtain evaluation scores.

Preferably, the intelligent template marking module includes template calculation and template marking; in the template calculation, the marked evaluation template image is identified and calculated to obtain the coordinates of the center point of the first evaluation box of the evaluation template, and the coordinate point information is saved; the first evaluation box center point coordinate set S _first is used to load a custom evaluation box control as a third evaluation box on the standard evaluation template image, and the evaluation box control in the third evaluation box is selected by dragging the mouse, and further marked to realize the association of all evaluation box controls in the third evaluation box with the evaluated object and evaluation indicators, and generate a template indicator library.

Preferably, the evaluation module includes an evaluator management submodule, an evaluation template management submodule, an evaluation subject maintenance submodule, an evaluation form printing submodule, an evaluation form upload identification submodule, an evaluation score submodule, an exception handling submodule and an evaluation result submodule.

Preferably, the evaluation template management submodule includes evaluation template creation, template indicator maintenance, template printing, and template copying functions. The evaluation template creation is used to manage the template content, including template basic information, paper size, number of pages, number of evaluation indicators, number of weights, number of evaluated objects per page, and upload the evaluation template file; the template indicator maintenance is used to create template indicators separately and set the scores according to the score setting rules, and the template indicator maintenance content is the indicator name and indicator score; the template printing is to use the template file to generate a marked evaluation template file and an empty evaluation template file, print, and further scan them as pictures to obtain marked evaluation template pictures and standard evaluation template pictures; the template copying realizes the reuse of templates when the same evaluation template is evaluated, and the copied template does not need to perform template coordinate calculation and template mapping; the weight refers to the sample weight, which is used to identify and distinguish the category attributes of the sample, so as to adjust the proportion or proportional distribution of samples in different groups participating in the collection of evaluation information when the total amount of evaluation samples is the same.

Preferably, the evaluation template management submodule includes evaluation template creation, template indicator maintenance, template printing, and template copying functions. The evaluation template creation is used to manage the template content, including template basic information, paper size, number of pages, number of evaluation indicators, number of weights, number of evaluated objects per page, and upload the evaluation template file; the template indicator maintenance is used to create template indicators separately and set the scores according to the score setting rules, and the template indicator maintenance content is the indicator name and indicator score; the template printing is to use the template file to generate a marked evaluation template file and an empty evaluation template file, print, and further scan them as pictures to obtain marked evaluation template pictures and standard evaluation template pictures; the template copying realizes the reuse of templates when the same evaluation template is evaluated, and the copied template does not need to perform template coordinate calculation and template mapping; the weight refers to the sample weight, which is used to identify and distinguish the category attributes of the sample, so as to adjust the proportion or proportional distribution of samples in different groups participating in the collection of evaluation information when the total amount of evaluation samples is the same.

Compared with the prior art, the present invention has the following beneficial effects:

The present invention provides an evaluation method based on image calibration, which uses the coordinate set of the center point of the first evaluation frame to automatically load the third evaluation frame through the template marking method, eliminating the tedious production process of filling in the template in the actual evaluation process, and realizing the horizontal and vertical batch marking of the evaluation template through the frame selection method, generating a template indicator library, and eliminating the need to manually mark the evaluation frame page by page when making the template, and to bind the evaluation indicators and personnel, thereby avoiding the problem of repeated rework and marking due to missed or wrong marking, and greatly improving the efficiency of template marking. In addition, a template marking can be supported for multiple repeated uses, realizing the printing of evaluation forms for all evaluation units under multiple evaluation themes of the same evaluation template and the recognition of filling results, overcoming the tedious operation of repeated marking when the same evaluation template is used by personnel from different units, and improving the work efficiency of the evaluators.

When recognizing the evaluation form, the template indicator library can quickly identify the evaluation form image after image standardization and obtain the evaluation score, which also greatly improves the recognition efficiency of the evaluation.

QR code technology is applied to the evaluation template. Each evaluation form has a unique identifier and does not support copying. Copying is invalid, which ensures the uniqueness, authenticity and anti-counterfeiting of the evaluation form results. It solves the problem of random copying and malicious filling in of evaluation forms in the actual evaluation process, which leads to deviations in evaluation results or even distortion of evaluations, and improves the fairness of the evaluation. At the same time, the use of QR code data information for rapid identification also improves the efficiency of filling in and identification.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a schematic diagram of a flow chart of an evaluation method based on image calibration according to an embodiment of the present invention;

FIG2 is a structural block diagram of an evaluation system based on image calibration according to an embodiment of the present invention;

FIG3 is a picture of a marked evaluation template according to an embodiment of the present invention;

FIG4 is a picture of a standard evaluation template according to an embodiment of the present invention;

FIG5 is a picture of an actual filling in of an evaluation subject evaluation form according to an embodiment of the present invention;

FIG6 is a schematic diagram of a template plotting flow chart according to an embodiment of the present invention;

FIG7 is a schematic diagram of a custom evaluation box control interface according to an embodiment of the present invention;

FIG8 is a schematic diagram of a horizontal frame selection according to an embodiment of the present invention;

FIG9 is a schematic diagram of a vertical frame selection according to an embodiment of the present invention;

FIG. 10 is a schematic diagram of a scoring process of an evaluation form image recognition according to an embodiment of the present invention.

DETAILED DESCRIPTION

The evaluation method and system of the present invention are applicable to many application fields that use two-dimensional tables as an information display method to achieve information collection, organization and analysis of evaluation data for specific purposes, such as various course evaluations in the education field, human resources personnel, training effects, interview evaluations, customer satisfaction in marketing, and other scenarios with large amounts of evaluation data.

As shown in FIG1 , an embodiment of the present invention discloses an evaluation method based on image calibration, comprising the following steps:

S1. Draw an evaluation template, create an image tag, and obtain an evaluation template file with an image tag;

S2. Based on the evaluation template file, obtain the marked evaluation template image and the standard evaluation template image;

S3, calculating the template coordinates based on the marked evaluation template image to obtain a coordinate set of the center point of the first evaluation frame;

S4. Based on the coordinate set of the center point of the first evaluation frame and taking the standard evaluation template image as the base map, the custom evaluation frame control is automatically loaded as the third evaluation frame, and the third evaluation frame is selected and marked to complete the template drawing and obtain the template indicator library;

S5. Maintain the evaluation theme, select the evaluation template, and print, distribute, and fill in all evaluation forms under the evaluation theme;

S6. Collect the filled evaluation forms, scan the evaluation forms, identify and calculate the evaluation form images, and obtain the evaluation scores.

The present invention is described below by taking a certain human resources theme evaluation of a certain enterprise as an example. The evaluation subjects of this theme evaluation are cadres and personnel of multiple subsidiaries, and the evaluation subjects of each subsidiary are more than 6 people. There are four evaluation indicators: personnel quality, job performance contribution, master-apprentice relationship, and style and image. Personnel evaluation is carried out around these four indicators, and evaluation scores are obtained. The scores are then analyzed for correlation to obtain the evaluation results. The specific steps include:

(1) Before the start of personnel evaluation, according to the specific evaluation system, the evaluation theme is planned, the evaluated unit, the evaluated person, the evaluation index, the index score, the evaluation weight, the evaluation time and other contents are set, the evaluation scope is determined, and the evaluation template is drawn on the computer using a report tool (such as the Grid++Report report designer, or the table generation function of the WPS office software) to create an image mark. The specific contents of the evaluation template include the evaluation title, QR code, evaluation unit, weight (sample weight, used to identify and distinguish sample categories), evaluation time, two-dimensional table (the horizontal row is the evaluated object and the vertical column is the evaluation index, or the vertical column is the evaluated object and the horizontal row is the evaluation index. In this embodiment, the horizontal row is the evaluated object and the vertical column is the evaluation index, and the table cell content is the fill-in area with different scores), instructions (i.e. notes), four corner positioning points, image marks and other evaluation-related contents; among which, the vertical column of the two-dimensional table is the evaluation index, the horizontal column is the evaluated object (name), and the dynamic Data, the content of the table cell is the score filling area, each score filling area is a rectangular area, and the area contains a score of 10-1; the image mark is a rectangular color block, which is dynamically configured on the evaluation template file and placed outside the two-dimensional table on the top and right side and corresponds to the score filling area in the evaluation template file. Specifically, it is located at the row mark above the first row of the two-dimensional table and the column mark on the right side of the last column of the table, and the row mark is the same as the width of the score filling area of the row with the most row scores displayed in the cells of the two-dimensional table or the same as the horizontal coordinate of the center point, and the column mark is the same as the height of the score filling area of the column with the most column scores displayed in the cells of the table or the same as the vertical coordinate of the center point. The image mark is combined with the row mark and the column mark to calculate the coordinates of the score filling area of the cells of the two-dimensional table, to assist in the coordinate calculation of the first evaluation box (all score filling areas in the marked evaluation template image); the evaluation subject, unit, template number, serial number, and weight are recorded in the QR code, which is dynamic data. Furthermore, in the evaluation system, upload the evaluation template file. The system uses the evaluation template file to generate a marked evaluation template file and a standard evaluation template file, which are connected to the printer. After printing, they are scanned to obtain the marked evaluation template image and the standard evaluation template image. The marked evaluation template image is shown in Figure 3, and the standard evaluation template image is shown in Figure 4 (the middle color block at the bottom is the page number mark). The scanning pixel is recommended to be above 600dpi.

(2) In the template calculation, select the evaluation template and calculate the coordinates of the center point of the first evaluation frame. During the calculation, the image of the marked evaluation template is firstly normalized to identify the QR code information in the image. The converted average ratio of the connected area of the positioning points on the image is _avp = _hp × _wp × _cp , where the length of the marked evaluation template image is _hp (px, i.e., pixel unit), the width is _wp (px, i.e., pixel unit), _cp is the percentage of the connected area of the four corner positioning points _Sp in the paper area _Sp0 , _cp = _Sp ÷ _Sp0 × 100%; the converted average ratio of the connected area of the marked evaluation template image and the image mark is calculated av _a = _hp × _wp × c _a , where c _a is the percentage of the connected area of the image mark S _a in the paper area _Sp0 , _cp = S _a ÷ _Sp0 × 100%; the calculation of the converted average ratio of the connected area of the positioning points is due to the inconsistency of the pixels of the scanned images. The area of the connected area of the positioning points is made consistent by converting the average ratio, and then the converted positioning points are used for perspective transformation, thereby ensuring the consistency of the coordinate systems of all images.

According to the resolution of the marked evaluation template image, the threshold range of the connected domain of the four corner positioning points is calculated as S _P ∈ (S _pmin , S _pmax ), where S _pmin = 0.8av _p , S _pmax = 1.2av _p ; the threshold range of the image mark connected domain is calculated as Sa _∈ (S _amin , S _amax ), where S _amin = 0.8av _a , S _amax = 1.2av _a ; the positioning points and image mark positions are determined by the identified ranges of S _P and _Sa ; for example, in the actual scanning process, the scanned image obtained may have problems such as tilt and non-standardization, and the standard coordinate system cannot be directly constructed. It is necessary to use the four corner positioning points to first correct the image, and then perform the coordinate system transformation to ensure that the coordinate systems of all images are consistent. As a preferred embodiment of this embodiment, the center point of the upper left corner of the image of the marked evaluation template is taken as the origin, and a two-dimensional coordinate system is constructed according to the plane where the image is located, wherein the horizontal axis of the two-dimensional coordinate system, i.e., the X-axis, extends from the origin of the image to the horizontal direction, and the vertical axis of the two-dimensional coordinate system, i.e., the Y-axis, extends from the origin of the paper to the vertical direction. The coordinates of the four corner positioning points of the marked evaluation template image are calculated as P _a1 (x _a1 ,y _a1 ), P _b1 (x _b1 ,y _b1 ), P _c1 (x _c1 ,y _c1 ), and P _d1 (x _d1 ,y _d1 ). The image is perspective transformed by the coordinates of the positioning points, so that the line connecting the upper left corner positioning point and the upper right corner positioning point and the line connecting the lower left corner and the lower right corner positioning points are parallel to the X-axis, and the line connecting the upper left corner positioning point and the center point of the lower left corner positioning point and the line connecting the upper right corner positioning point and the center point of the lower right corner positioning point are parallel to the Y-axis, so that the image that is not scanned correctly can be corrected. After correction, the image is cropped into a rectangular area and transformed into a new coordinate system, which facilitates the recognition and calculation of the image in the positioning point area and ensures the consistency of the two-dimensional coordinate system during recognition.

The X-axis and Y-axis coordinates of the four corner positioning points are sorted. As a preferred embodiment of this embodiment, a point with the smallest X-axis and Y-axis coordinate values is recorded as the origin of the new coordinate system O(0,0). After the coordinate system transformation is completed, the coordinates of the four corner positioning points of the two-dimensional table and all image marks are calculated, wherein the coordinates of the four corner positioning points of the two-dimensional table are recorded as E _a1 (x _ea1 ,y _ea1 ), E _b1 (x _eb1 ,y _eb1 ), E _c1 (x _ec1 ,y _ec1 ), E _d1 (x _ed1 ,y _ed1 ), and the table is subjected to a coordinate perspective transformation according to the coordinates to achieve correction of the table. In this embodiment, as a preferred embodiment, a point E _b1 (x _eb1 ,y _eb1 ) with the largest horizontal coordinate value and the smallest vertical coordinate value among the four corner positioning points of the two-dimensional table is used as a reference to identify the position of the image mark. That is, the upper right corner point of the four corner positioning points of the two-dimensional table is used as a reference for easy search and identification, and compared with the horizontal and vertical coordinate values of the image mark to clearly distinguish and identify the position and coordinates of the image mark above and to the right of the two-dimensional table, and then calculate the coordinate set of the center point of the first evaluation frame.

The position of the image mark is identified based on the coordinates of E _b1 (x _eb1 ,y _eb1 ). When the horizontal coordinate x _a of the center point of a certain image mark differs from x _eb1 by more than 10px, that is, x _a -x _eb1 >10px, the point is located on the right side of the two-dimensional table, and the vertical coordinate of the point is marked as Yi _. When the vertical coordinate y _b of the center point of a certain image mark differs from y _eb1 by more than 10px and within 200px, that is, 10px <y _eb1 -y _a <200px, the image mark of the point is located above the two-dimensional table, and the horizontal coordinate of the point is marked as _Xi . The coordinate values _Ai ( _Xi , _Ya ) and _Ri ( _Xb , _Yi ) of the upper and right image marks are calculated, and the horizontal coordinate _Xi of the upper image mark is combined with the vertical coordinate value _Yi of the right image mark to calculate the coordinates _Ci ( _Xi , Yi ₎ of the center points of all evaluation frames in the marked evaluation template image. ), and obtain the coordinate set of the center point of the first evaluation frame S _first ={C ₁ (X ₁ ,Y ₁ ),C ₂ (X ₂ ,Y ₂ ),…,C _n (X _n ,Y _n )}. The image standardization process is to perform grayscale conversion, binary segmentation and opening operation on the image.

(3) In the template marking of the intelligent template marking module, select the evaluation template, automatically load the third evaluation frame on the standard evaluation template image, and use the mouse to select and mark the third evaluation frame in the interface to complete the template marking. The specific steps are as shown in Figure 6, the template marking process of this embodiment:

In template marking, select the assessment template to be marked and enter the template marking interface. Preferably, the marking interface automatically loads the standard assessment template image corresponding to the assessment template, performs image standardization on it, and also masks the QR code and the words of the assessment unit in the image to use it as the base map for visual display of the marking.

According to the information of the center point coordinate set S _first of the first evaluation frame, the custom evaluation frame control is dynamically and automatically loaded as the third evaluation frame. The custom evaluation frame control automatically generates a score filling area control with the same size and style as the standard evaluation template image in the interface according to S _first . As the third evaluation frame, it is displayed in blue by default. As shown in Figure 7, the interface diagram of the custom evaluation frame control of this embodiment can be selected by the mouse. The properties of the custom evaluation frame control include score, score color, score font size, drawing line width, height, drawing line color, belonging indicator, belonging object, unique identifier, and center point coordinates.

On the interface, the third evaluation box can be selected through the mouse selection function. Preferably, the selection function uses the mouse dragging and brush drawing functions to display the user's dragging area on the screen, calculate the controls in all the third evaluation boxes in the dragging area, and change their selected colors to red, and enter the "【Indicator】Batch Labeling" interface at the same time. The labeled content includes evaluation objects and evaluation indicators. The "【Indicator】Batch Labeling" realizes the associated storage of all selected controls in the third evaluation box with the evaluation objects and evaluation indicators by selecting and binding the selection area with the corresponding evaluation object number and evaluation indicator number.

The selection is done by holding down the left mouse button and dragging. The drag selection area is displayed with a black dotted line on the interface. When the left mouse button is released, the third evaluation box control in the selection area changes from blue to red, and the black dotted line disappears. It supports multiple selection rules, such as selection by evaluation indicator, selection by evaluation object, selection by filling in the area control by a single score, and marking. After marking, the evaluation box automatically records the marking information according to the properties of the custom evaluation box control.

The selection by evaluation index is a horizontal selection by a rectangular frame, such as a1 and a2 in Figure 8. For all score filling area controls in the area, the number of indicator boxes and the number of indicator scores in the selection range are compared from left to right or from right to left according to the indicator score setting rules. After the comparison is correct, the score attributes and indicator attributes are set according to the indicator score setting rules.

The selection by evaluation object is based on the evaluation template layout, and is performed vertically through a rectangle, such as b1 and b2 in Figure 9, from top to bottom or from bottom to top. All score filling area controls in the rectangular area correspond to the same evaluation object, and the evaluation object number is set.

To select the box by filling in a single score area control, double-click the score fill-in area control in the third evaluation box with the mouse. You can bind the indicator information, evaluation object information and score information through the [Indicator] single annotation. After saving, only the single third evaluation box control that you double-clicked with the mouse will be associated and stored with the selected information.

The annotation is to bind the evaluation indicators and the evaluation objects through batch annotation or single annotation after the box selection is completed. Preferably, all the evaluation box controls in the third evaluation box are associated and bound with the evaluation objects and evaluation indicators through horizontal and vertical batch annotation and stored in the database to obtain the template indicator library of the current evaluation template.

(4) After completing the above preparations, enter the personnel evaluation step. The specific steps include: maintaining the evaluation topic in the evaluation module, selecting the evaluation template, setting the evaluation unit and personnel for the evaluation template; setting the start and end time of the evaluation topic; connecting to the printer, printing the personnel evaluation form under the topic, and distributing and filling it out.

(5) After the evaluation form is filled in, it is collected uniformly, and the computer is connected to the scanner. The evaluation form is scanned by the scanner to obtain an evaluation form image. Figure 5 is an actual filled-in picture of an evaluation form for an evaluation subject in this embodiment. The system processes and recognizes the evaluation form image to obtain the evaluation score. When performing image recognition of the personnel evaluation form, it supports batch scanning of evaluation form papers to obtain images and then upload them to the system for recognition. It also supports batch scanning of evaluation forms while the scanner is scanning and the system is recognizing them. The specific steps for processing and recognizing the evaluation form image are as follows:

The system recognizes the QR code in the current evaluation form image, obtains the evaluation template information, and calls out the template indicator library of the evaluation template.

The average conversion ratio of the connected area of the positioning points on the paper and the current evaluation form image is calculated as _avp1 = _hp1 × _wp1 × _cp1 , where the length of the image is _hp1 (px), the width is _wp1 (px), _cP1 is the percentage of the connected area of the positioning points _Sp1 to the paper area _{Sp10 cp1} = _Sp1 ÷ _Sp10 ×100%; the threshold range of the connected domain of the four corner positioning points is calculated as _Sp1∈ ( _Sp1min , _Sp1max ), where _Sp1min = _0.8avp1 , _Sp1min = _1.2avp1 ; for example, as a preferred embodiment of the present invention, the center point of the upper left corner of the evaluation form image is taken as the origin, and a two-dimensional coordinate system is constructed according to the plane where the image is located, wherein the abscissa axis of the two-dimensional coordinate system, i.e., the X-axis, extends from the origin of the image to the horizontal direction, and the ordinate axis of the two-dimensional coordinate system, i.e., the Y-axis, extends from the origin of the paper to the vertical direction, and the coordinates of the four positioning points on the current evaluation form image are calculated as _P1a1 ( _xa1 ,y _a1 ）,P1 _b1 (x _b1 ,y _b1 ), P1 _c1 (x _c1 ,y _c1 ),P1 _d1 (x _d1 ,y _d1 ), the image is perspective transformed by the coordinates of the positioning points to correct the image that is not scanned correctly. The X-axis and Y-axis coordinates of the four positioning points are sorted. As a preferred embodiment of this embodiment, the point with the smallest X-axis and Y-axis coordinate values is recorded as the origin of the new coordinate system O1 (0,0), and a rectangular cut is performed with the coordinates of the four positioning points in the current evaluation form image to calculate. Then the coordinate system is the same as the coordinate system in the template indicator library. The query range is expanded by a certain number of pixels using the coordinates of the center points of all the filled points in the template indicator library. The coordinates of the starting point and the ending point of the query range are calculated, and the range is cut into a rectangle. The pixels in the area are calculated, and the percentage of pixels is used to determine whether it is a filled point. If it is a filled point, the corresponding indicators and personnel scores of all the filled points in the current evaluation form image are obtained based on the coordinates of the center point of the filled point and the content in the template indicator library.

According to the current evaluation form image recognition situation, for evaluation form images with abnormal filling, such as over-filling, missing filling, etc., manual processing will be performed item by item. The system supports manual adjustment of saved scores, and also supports discarding and recalculation.

(6) Through steps S1 to S6 in this embodiment, the evaluation results of all participants are obtained, and the evaluation results are analyzed for correlation between the total score and each indicator sub-item, between different indicator sub-items, and between the scores of people with different weights, to obtain the evaluation results.

As shown in FIG. 2 , an embodiment of the present invention further discloses an evaluation system based on image calibration, in which S1 to S6 in the above method are implemented, and the system includes:

Template drawing module, used for drawing evaluation templates and dynamically configuring template content, and obtaining evaluation template files with image tags and QR codes;

Intelligent template plotting module, used to calculate template coordinates, use the evaluation frame control to dynamically generate the evaluation frame and plot the template, and generate a template indicator library;

The evaluation module is used for the management of evaluation templates, evaluation template indicators, evaluation topics, and evaluated objects, and further completes the printing of evaluation forms for all evaluated objects for evaluation; collects, scans, and identifies filled evaluation forms, performs identification and calculation based on the template indicator library, obtains evaluation scores, and analyzes the scores to obtain evaluation results.

According to the actual evaluation system, the evaluation template is drawn and the template content is dynamically configured. The evaluation template file with image tags is obtained through the template drawing module. The file includes at least image tags, QR codes, evaluation topics, four corner positioning points, two-dimensional tables and other information. The image tags support various regular shape color blocks or shape connected areas.

The evaluation template calculation and template marking are performed in the intelligent template marking module. When calculating the template, the marked evaluation template image is identified and calculated to obtain the coordinate set of the center point of the first evaluation frame and save the coordinate point information; the third evaluation frame is automatically loaded using the coordinate set of the center point of the first evaluation frame, and the third evaluation frame is selected by dragging the mouse to select it, and the selected area is batch marked to associate the third evaluation frame with the evaluated object and the evaluation indicator, and generate a template indicator library.

The assessment module includes the assessor management submodule, the assessment template management submodule, the assessment subject maintenance submodule, the assessment form printing submodule, the assessment form upload identification submodule, the assessment score submodule, the exception handling submodule and the assessment result submodule; before maintaining the assessment subject, the assessed object and unit information are maintained in the assessor management submodule.

For example:

Unit: It is to manage the information of multiple subsidiaries or departments under the company according to the actual organizational structure of the assessment unit.

Evaluated object: The personnel information is maintained according to the evaluation unit. When maintaining personnel, the position information needs to be maintained in advance.

Positions: deputy general manager, middle-level deputy position, middle-level full-time position, high-level skilled talents, high-level scientific and technological talents, special staff, leaders, ministers, etc.

In the template management submodule, the template is created, the template indicators are maintained, the template is printed, and the template is copied. When creating a template, the template content is managed, such as the basic information of the template, the paper size (in this embodiment, A4 paper is selected to print and produce the evaluation form), the number of pages, the number of evaluation indicators (in this embodiment, 4 indicators are selected for evaluation), the weight number (in this embodiment, 6 types of sample weights are selected for evaluation, that is, A, B, C, D, E, F, a total of 6 types of weighted personnel, setting the sample weight, the number of samples can be allocated and adjusted by category, and when the total number of evaluation samples is fixed, it is convenient to conduct the correlation analysis between the sample category and the evaluation object and the evaluation indicator, so that the evaluation results serve the evaluation goals), the number of evaluated objects per page (in this embodiment, each evaluation form of the evaluation template can be evaluated for up to 6 objects), upload the marked evaluation template file; further maintain the four indicator information, create the indicators separately and set all the values, mainly including the indicator name and indicator score. For example, if the indicator score of the personnel quality indicator is 1-10, then when setting it, according to the score setting rules, 10 filling areas are set in the cells of the two-dimensional evaluation table according to the score, and the indicator scores of 1-10 for the quality of the 6 evaluated objects are assigned in sequence, which can be maintained as follows:

10,,9,10,,9,10,,9,10,,9,10,,9,10,,9,8,,7,8,,7,8,,7,8,,7, 8,,7,8,,7,6,5,4,6,5,4,6,5,4,6,5,4,6,5,4,6,5,4,3,2, 1,3,2,1,3,2,1,3,2,1,3,2,1,3,2,1.

Maintain the four evaluation indicators one by one according to the score setting rules.

Utilize the evaluation template file with image marking to generate marked evaluation template file and standard evaluation template file in the system, print and scan to obtain marked evaluation template image and standard evaluation template image, which are used for template coordinate calculation to obtain the coordinate set of the center point of the first evaluation frame; according to the coordinate set of the center point of the first evaluation frame, dynamically load the custom evaluation frame control, and perform template marking to generate a template indicator library.

Template copying is mainly used to reuse templates when the same template is used for different evaluation topics. The copied template does not need to repeat the calculation and plotting of the template, which can greatly improve the work efficiency of the evaluation.

The evaluation theme maintenance submodule implements the creation of evaluation themes. When creating, select the evaluation unit, evaluation start time, evaluation end time, and evaluation template, select the unit corresponding to the evaluation template, and perform personnel maintenance. One evaluation theme can be used for evaluation in multiple units. Similarly, one evaluation theme can select multiple evaluation templates for evaluation, and one evaluation template can be used for evaluation in multiple units. When maintaining the evaluated objects of the theme, the evaluated objects of different units under different evaluation templates can be selected one by one. The system automatically generates evaluation tables for all evaluated objects under different evaluation templates of all evaluation units under the evaluation theme.

The evaluation form printing submodule prints evaluation forms of different evaluation templates for all units under the evaluation theme. It supports custom printing of copies with different weights, and also supports printing of a certain evaluation form. Each evaluation form is unique, and the QR code on each evaluation form is a unique identifier. The evaluation form supports multiple repeated printings, but does not support copying. The information recorded in the QR code includes template number, evaluation theme, serial number, weight, and page number information.

After the printed evaluation form is distributed and filled in, the identification submodule is uploaded to the evaluation form for scanning and identification, and the score is calculated to obtain the evaluation score. The step flow is shown in Figure 10. The evaluation image recognition and scoring process of this embodiment. The system is connected to the scanner to perform batch scanning of the filled evaluation form, and the scanned evaluation form is subjected to image standardization processing before comparison and identification, and score calculation is performed; for the evaluation form with no abnormality, the score result is saved according to the QR code information in each evaluation form; for the evaluation form with abnormality, the abnormal result can be processed.

In the exception handling submodule, the recognition results are used to save the evaluation form images that are filled in manually and cannot be recognized in the exception library, and the scanned images of the evaluation form and the exception type are displayed. Select the recognized abnormal data, and manually process the recognized abnormal data according to the exception type. After the manual adjustment is completed, the score results are saved; support recalculation and discarding of abnormal images. The exception types include recognition exceptions, evaluation subject does not exist, evaluation template does not exist, template indicator details do not exist, template score details do not exist, evaluated personnel details do not exist, the number of scores is inconsistent with the set number, the score is not recognized, the QR code recognition exception, the score does not match, the evaluated object does not match, the number of scores is inconsistent with the set number, etc.

In the evaluation result submodule, a correlation analysis is performed based on the evaluation scores and using the pre-set weight coefficients. The main focus is on the correlation analysis between the total score and each indicator sub-item, between different indicator sub-items, and between the scoring results of people with different weights, and ultimately the personnel evaluation results under the current evaluation topic are obtained.

To run the above evaluation method and system, the computer, printer and scanning equipment provided by the present invention are configured as follows:

computer:

Operating system: Windows 7 and above;

Memory: 16GB and above;

Hard disk: 500G and above.

printer:

Type: Black and white laser printer;

Maximum printing format: A4;

Maximum resolution: 600×600dpi;

Printing method: laser printing.

Scanner:

Type: Sheet-fed;

Scanning light source: LED;

Scanning modes: color, grayscale, black and white;

Optical resolution: 600×1200dpi;

Scanning range: 245×356mm;

Output formats: support bmp, jpg, multi-page tiff, multi-page pdf formats.

The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims (10)

1. An evaluation method based on image calibration, characterized in that the method comprises the following steps: S1. Draw an evaluation template, create an image tag, and obtain an evaluation template file with an image tag; S2. Based on the evaluation template file, obtain the marked evaluation template image and the standard evaluation template image; S3, calculating the template coordinates based on the marked evaluation template image to obtain a coordinate set of the center point of the first evaluation frame; S4. Based on the coordinate set of the center point of the first evaluation frame and taking the standard evaluation template image as the base map, the custom evaluation frame control is automatically loaded, the template is marked, and the template indicator library is obtained; S5. Maintain the evaluation theme, select the evaluation template, print, distribute and fill in the evaluation form; S6. Collect the filled evaluation form, scan the evaluation form image, identify the filled score, and obtain the evaluation result. 2. The evaluation method as described in claim 1 is characterized in that, in step S1, the evaluation template file with image marks includes an evaluation title, a QR code, an evaluation unit, a weight, a two-dimensional table, instructions, four corner positioning points and an image mark; the image mark is a connected domain color block that is dynamically configured on the evaluation template and is placed outside the two-dimensional table and corresponds to the filled area in the evaluation template, and is used as a mark to assist in the coordinate calculation of all filled areas in the first evaluation box; the first evaluation box is a collection of all filled areas in the two-dimensional table in the marked evaluation template image. 3. The evaluation method as described in claim 2 is characterized in that in step S3, the template coordinate calculation includes image standardization processing, recognition and calculation of the evaluation template picture with image marks, determining the positioning points and image mark positions and correcting the image; the image standardization processing is to convert the image into grayscale, binarize the image and perform opening operations. 4. The evaluation method as described in claim 1 is characterized in that in step S4, the template marking is to use a mouse or a touch device to select and mark the third evaluation box. 5. The evaluation method as described in claim 1 is characterized in that, in step S6, the evaluation form scanning and recognition includes image standardization processing of the evaluation form image, identifying the filled area and calculating the score according to the template indicator library. 6. An evaluation system based on image calibration, characterized by comprising: The template drawing module is used to draw the evaluation template and dynamically configure the template content to obtain the evaluation template file with image tags; An intelligent template marking module is used to calculate the template coordinates to obtain the coordinate set S _first of the center point of the first evaluation frame, and based on S _first , load a custom evaluation frame control as the third evaluation frame on the standard evaluation template image, select and mark the frame, complete the template marking, and generate a template indicator library; The evaluation module is used to manage evaluation templates, evaluation template indicators, evaluation topics, and evaluated objects, complete the printing of evaluation forms for all evaluated objects, and conduct evaluations; collect, scan, and identify filled-in evaluation form papers, and calculate based on the template indicator library to obtain evaluation scores. 7. The evaluation system as described in claim 6 is characterized in that the intelligent template marking module includes template calculation and template marking; in the template calculation, the marked evaluation template image is identified and calculated to obtain the coordinates of the center point of the first evaluation box of the evaluation template, and the coordinate point information is saved; the first evaluation box center point coordinate set S _first is used to load a custom evaluation box control on the standard evaluation template image as a third evaluation box, and the evaluation box controls in the third evaluation box are marked by dragging the mouse to select the box, so as to realize the association of all evaluation box controls in the third evaluation box with the evaluated object and the evaluation indicators, and generate a template indicator library. 8. The evaluation system as described in claim 6 is characterized in that the evaluation module includes an evaluation personnel management sub-module, an evaluation template management sub-module, an evaluation subject maintenance sub-module, an evaluation form printing sub-module, an evaluation form upload identification sub-module, an evaluation score sub-module, an exception handling sub-module and an evaluation result sub-module. 9. The evaluation system as described in claim 8 is characterized in that the evaluation template management submodule includes evaluation template creation, template indicator maintenance, template printing, and template copying. 10. The evaluation system as described in claim 8 is characterized in that the evaluation form upload identification submodule is connected to a scanner to realize batch scanning of filled evaluation forms, obtain evaluation form images, process and identify the evaluation form images, and calculate scores.