TWI642021B - Quick and easy optical identification point menu editing system - Google Patents

Abstract

The invention relates to a quick and easy optical identification point menu editing system, which mainly comprises: a customized blank point menu, an image processing module, an image recognition module, an image correction module, an editing module, and the same material. The library is composed of; the image file of the customized blank order menu can be prepared, or the image capture module can be used to capture the customized blank order menu through the camera lens, and archived for use by the image processing module. After a series of effective related treatments, the restaurant operators can design and edit their own, and simply complete the exclusive and easy-to-use optical identification menu of their own restaurant, without having to entrust the professionals to carry out related menu changes. Time and effort can reduce costs.

Description

Quick and easy optical identification point menu editing system

The invention relates to a simple editing system for a quick and easy optical identification point menu, in particular to an effective practical function, which allows the catering industry to design and edit by itself, and simply complete the restaurant's exclusive quick and easy optical identification menu.

Although the patent publication number 201220223 "reading ordering system" can be used to solve the problem of ordering meals in the traditional catering industry, it must be used as an input when the implementation is performed, and the design on the menu is also compared. Complex, in addition to the corner to have a basic positioning mark, the menu must also be added to the entire block of the correction map group, the overall design is more complex, high manufacturing costs, poor practicality is still not perfect.

To this end, the inventors have previously designed a quick and easy optical identification menu for the patent application No. 106103118, which is mainly provided with three positioning marks on the three corners of the order menu as the origin of the ordering, The X-axis and Y-axis reference coordinate values, and a plurality of selection frames on the order menu, each of the selection frames corresponding to the three positioning marks having a relative coordinate value, and the selection frame also has a corresponding coordinate frame Selection of content; By using this fast and easy optical identification menu, the effective design can be The information that the customer has to fill out when ordering is formatted in advance, and replaced by the selection frame of the fixed coordinate point. The printed order list is simple and beautiful, and the two sides do not leave the entire compartment like the custom structure. By arranging the obscured correction map set, it is no longer necessary to use a dedicated scanning device. As long as the simple optical image is read and recognized, accurate ordering content can be obtained, which is convenient for use as a checkout and subsequent feeding. Because the application is quite convenient, it is not necessary to input the customer's ordering content one by one by the service personnel, so the service personnel can shorten the ordering time of the food input one by one, and effectively avoid the probability of human input errors, and greatly improve the service quality of the ordering meal. .

However, the conventional POS system of the computer terminal cash register POS system does not have such a simple editing system for the quick and easy optical identification menu. Therefore, it is still necessary to entrust the image processing developer to complete the blank order menu. The associated database, when facing the restaurant operation needs to change the menu, you have to do another editing of the new menu in the image department, because these tasks need to be performed by professional engineers, may also delay to the original reservation replacement menu Time, causing inconvenience in the operation of the restaurant.

In view of the above, the main object of the present invention is to provide a quick and easy optical identification point menu editing system, and the first embodiment mainly includes: a customized blank point menu, an image processing module, and an image recognition module. a group, an image correction module, an editing module and the same database, wherein the customized blank order menu has a corresponding image file; preferably The image file of the customized blank order menu can be prepared in advance. The format of the image file can be color or gray scale. The image capture module can also be used to capture the customized blank order menu through the camera lens. And storing the image file for use; the image processing module converts the image file of the blank order menu from a color format to a gray scale format; the image recognition module identifies the image file of the order image The positioning mark and the selection frame store the average gray scale value and the relative coordinates of the positioning mark and the selection frame in the blank point menu; the image correction module performs deviation correction on the image of the order image file to obtain the regularity The origin and the coordinate axis, and standardize the coordinates of the selection frame; the editing module provides the standardized coordinate of the edited selection frame and the average grayscale difference thereof for the restaurant manager to subsequently input the associated database. The corresponding meaning of the selected picture frame, and the sample database stores the standardized coordinate value, the average gray level difference, and the corresponding meaning of the selection frame of the selected sampling frame; The invention discloses a quick and easy optical identification point menu editing system, and the second embodiment mainly comprises: two customized order menus, an image recognition module, an image correction module, an editing module and the same material. The library consists of two custom-made menus, one for the original box, and one for the selected menu, and the two menus have corresponding Image file The image recognition module identifies the positioning marks on the image files of the two order menus, and sets a frame with a difference in pixel values compared to the contents of the image files of the two order menus. The image correction module performs deviation correction on the maps of the two order menu image files to obtain the normalized origin and coordinate axes, and normalizes the coordinates of the selection frame. The editing module provides a standardized coordinate of the selection frame and an average pixel value thereof for the restaurant manager to subsequently input the related database to link the corresponding meaning of the selected frame, and the sample database The standardization coordinate value, the average pixel value, and the corresponding meaning of the selection frame of the selected sampling frame are stored; thereby, the food and beverage industry can design and edit the decoration, and simply complete the customization of the restaurant. Order a meal without having to entrust a professional to make a related menu change. The entire operation is relatively time-saving and labor-saving in the design of the menu content, which can reduce the cost and shorten the time for changing the menu, so as to improve the operational efficiency of the restaurant.

10‧‧‧Image capture module

100‧‧‧ order menu

100’‧‧‧ order menu

101‧‧‧ Positioning Mark

102‧‧‧Selection frame

110‧‧‧image file

110’‧‧·image file

20‧‧‧Image Processing Module

30‧‧‧Image recognition module

40‧‧‧Image Correction Module

50‧‧‧editing module

60‧‧‧ sample database

1 is a block diagram of a first embodiment of a quick and easy optical identification point menu editing system of the present invention.

FIG. 2 is a diagram showing an example of editing application of the quick and easy optical identification point menu according to the first embodiment of the present invention.

Figure 3 is a schematic diagram of editing of the positioning mark of the present invention.

Fig. 4 is a schematic view showing the sampling of binarization of the selection frame of the present invention.

FIG. 5 is a view showing an application example of image recognition according to the first embodiment of the present invention.

Figure 6 is a block diagram showing a second embodiment of the present invention.

Figure 7 is a schematic diagram showing a blank menu box for a meal order according to a second embodiment of the present invention.

Figure 8 is a schematic view showing the ordering of the ordering list in the second embodiment of the present invention.

FIG. 9 is a view showing an application example of image recognition according to a second embodiment of the present invention.

In order to facilitate the understanding of the contents of the present invention and the achievable effects, the specific embodiments are listed in conjunction with the drawings, and the detailed description is as follows: Referring to Figures 1 to 5, the present invention provides a quick and easy optical identification ordering meal. The first embodiment of the single editing system includes: a customized blank order menu 100, an image processing module 20, an image recognition module 30, an image correction module 40, an editing module 50, and the like. The database is composed of the database 60, wherein the customized blank order menu 100 has a corresponding image file 110, and the format of the image file 110 can be color or gray scale; when implemented, an image capture module can also be used. 10, through the photographic lens to capture the customized blank order menu 100, and then stored as an image file 110 for the image processing module 20; the image processing module 20, using the blank point menu 100 image The file 110 is converted from a color format to a grayscale format to shorten subsequent operations and processing time. The image recognition module 30 is configured to identify the positioning mark 101 and the selection frame 102 on the order image file 110, and store blank spots. The average grayscale value of the selection frame 102 in the menu 100; the average grayscale value of the positioning mark 101 and the selection frame 102 in the blank order menu 100, and the coordinate value of the selection frame 102 relative to the positioning mark 101; The image correction module 40 performs offset correction on the map of the order image file 110 to obtain the normalized origin and coordinate axes, and calculates the coordinates of the standardized selection frame 102; the editing module 50 provides editing and selection. The normalized coordinates of the block 102 and the average grayscale difference are used by the restaurant manager to subsequently input the associated database for linking the corresponding meanings of the selected frame 102, and the sample database 60 stores the aforementioned The calibration sampling frame 102 normalizes the coordinate value, the average grayscale difference, and the corresponding meaning of the selection frame 102; thereby, the food and beverage industry can design and edit by itself, and simply complete the self-customized instant and easy optical identification of the restaurant. Menus, without having to entrust professionals associated with the menu changes work. The entire operation is relatively time-saving and labor-saving in the design of the menu content, which can reduce the cost and shorten the time for changing the menu, so as to improve the operational efficiency of the restaurant.

More specifically, the image capturing module 10 of the present invention can capture the image of the blank order menu 100 through the CCD or CIS camera lens, and save the image. The image file 110 may be in the form of color or grayscale. Or, through the image processing module 20, the image file 110 is converted from a color format to a grayscale format to shorten subsequent operations and processing time.

Since the common menus in the market may be color and colorful images, first changing the image from color to gray scale can greatly reduce the size of the image file and speed up the processing for subsequent processing. RGB (red, green, blue) is The three primary colors of color, the combination of the three primary colors of different intensities can present all colors. In the RGB color model, the completely saturated light is combined into white (100% R, 100% G, 100% B), and the white complementary color is black (0 % R,0% G,0% B), the same intensity of red, green and blue combined into a middletone gray scale, for color images, the color to gray scale can reduce the amount of data to the original three-thirds First, the reduction in the amount of data can increase the processing speed and reduce the storage space without affecting the accuracy of data interpretation. The basic operation formula of the preferred implementation is as follows: the grayscale value Gray=(Red Red+Green Green+Blue Blue)/3, but the actual limit is not limited thereto.

Although the above description shows the gray scale as an example, in fact, the same identification purpose can be achieved by the color RGB method, but the recognition takes a little more time, for example, the image value = (Red+Green+Blue)/3, or even Use different weighting methods to calculate, such as pixel value = .299 * Red + 0.587 * Green + 0.114 * Blue.

With regard to the image recognition module 30, the positioning mark 101 and the selection frame 102 are mainly recognized in the following manner, and the blank point menu 100 is selected for selection. The average grayscale value of block 102 and the coordinate value relative to the anchor mark 101:

a. As shown in Fig. 4, with image binarization, the grayscale image is divided into two gray values, that is, a threshold value is set, when the image grayscale is greater than the threshold. The value is such that its grayscale is a bright spot of 225, whereas if the grayscale is less than the threshold value, the grayscale is a defect of 25. The gray scale statistics of most binarized images have a bimodal characteristic. The statistical principle can be used to find the optimal threshold setting. When the grayscale value of the image is less than the defined threshold (Threshole), it is selected. If the grayscale value is greater than the threshold value, it is blank.

b. Identifying the positioning mark 101: using the pixel grayscale value obtained above, extracting the adjacent pixel but the grayscale has a large change area, and comparing the geometrical figure of the area with the built-in standard positioning mark 101, such as the geometric figure The fixed ratio of the built-in standard positioning mark 101 is taken as the positioning mark 101, and the positioning mark 101 in the upper left corner is set as the origin, and the gray scale value of the black body portion is recorded as the reference value, which can be The scale of the "horizontal X-axis" is set to the origin X-axis length, and the "horizontal Y-axis" scale is set to the origin Y-axis length. Then, the selection frame 102 is relative to the positioning mark 101, and its position can generate a two-dimensional coordinate system (X, Y), and the average gray level value of the selection frame 102 can generate an average gray-scale difference value for the reference value. As shown in FIG. 3, it is an example of the geometry in which the standard positioning mark 101 is established in the image recognition module 30.

c. Identification selection frame 102: using the pixel grayscale values obtained above, extracting adjacent pixels but the grayscale has a large change region, and comparing the geometry of the region with the built-in standard selection frame 102, such as geometry The graphic conforms to the built-in standard selection box 1 For a fixed ratio of 02, the geometry is captured as a selection box 102, and its coordinate value to the origin and its average grayscale value are recorded.

Further, the image correction module 40 is mainly used to correct the origin and the coordinate axis and confirm the normalized coordinates of the selection frame 102. This is to avoid the distortion of the image caused by factors such as light or placement position when the order list of the manufacturer is prevented from being read by the image capturing module 10. This phenomenon causes an error when comparing the selection frame 102, so the image is taken. After the completion, the digital image processing technology is used to correct the coordinate image and compare the coordinate values of the positioning mark 101 and the built-in standard positioning mark 101, and use mathematical processing such as scaling and rotation to make the coordinate values of the three positioning marks 101 and The coordinate values of the built-in standard positioning mark 101 (such as the sample order list, the length of the X-axis is 33 and the length of the Y-axis is 51) coincide, and the conversion mathematical formula between the two is recorded, for example: X'=aX+bY

Y’=cX+dY

The coordinates of each of the selection frames 102 are converted into coordinates by the above-described conversion formula.

Referring to FIG. 5, the data collected by the image recognition module 30 and edited by the image correction module 40 may be pre-recorded into the sample database 60: the converted frame 102 is normalized to coordinates and gray scales. The difference is used by the restaurant's management staff to enter the associated database, ie (the standardized coordinate value of the sample box and the average grayscale difference) and the relevance of the sample box (specific meaning or related information of the sample box), The content that allows the customer to order the meal can be read and recognized with the restaurant computer. The terminal cash register POS system is corresponding to facilitate the correct ordering, checkout and subsequent meal.

Please refer to FIG. 2 again, which is an application example of the editing of the quick and easy optical identification point menu of the present invention, wherein the positioning marks 101 located at three corners can form a coordinate system and have a selection of meals and related meal information. Area, as well as instructions for using artwork and order menus. The dashed grids in the figures and their corresponding scales are not required for practical application, only to facilitate the description of the principles of the invention and its specific implementation.

In this embodiment, the scale is defined by the length and width of the black mark of the positioning mark 101, and the positioning mark 101, the shape of the mark (the inner circle is all black) can be expressed by the format of the third figure (in practical application, the unit) The length can be defined by other geometric shapes. In this example, the length of the X-axis is 33 units, the length of the Y-axis is 51 units, and each of the selection frames 102 is exactly a square of unit length (the X-axis in actual application) The length with the Y axis, and the length and width of the selection frame 102 may be non-integer).

Please refer to Figures 2 and 4. In the example, each of the selection frames 102 will be in this coordinate system, the coordinate points (X1 & X2) around the X axis, and the coordinate points above and below the Y axis (Y1 & Y2) is defined, (in practice, X1, X2, Y1, Y2 can also be non-integer): so each of the selection boxes 102 is on the order list, not only the corresponding coordinate points (X1, X2, Y1) , Y2), each also has its specific meaning or related content, as shown in the selection box 102 (X1=16, X2=17, Y1=10, Y2=11), the corresponding information For "meal name = sweet and sour squid and unit price = 380 and quantity = 1".

Therefore, when the image capturing module 10 is used to capture the blank order menu 100, the gray positioning data of the image has different grayscale values from the blank marking mark 101, such as the origin. The grayscale value of the positioning mark 101 is a reference, and the grayscale value difference between each blank selection frame 102 and the "origin positioning marker 101" in the blank order menu 100 is calculated, and the grayscale difference of FIG. 4 is obtained. The value is the result of binarization.

In this way, after the user selects the quick and easy optical identification menu, the selected grayscale value difference between the original and the original positioning marker 101 will be less than the selected one. The blanking box 102 is low, and the change of the difference can be used as a criterion for determining whether the status of the selection box 102 on the customer order list is "unselected" or "selected".

In this way, the sample database 60 can be easily established to link the tabs 102 on the blank menu 100 to normalize the coordinates, the average grayscale difference, and the specific meanings of the corresponding selections, such as the table number, the number of people, Information such as meal name, amount and quantity. The simple design of the exclusive menu that meets the needs of the home provides customers with convenient services such as ordering, checkout and follow-up meals.

The above is only a simple illustration, but it is not limited to this. For example, the positioning mark 101 can be composed of □, ○, ◎, △, ◇, × in addition to the shape of the back (the inner ring is all black). , + and ♁ and any of these geometric symbols. In addition to the square shape, the selection frame 102 may be one of a circle, a rectangle, and a long strip, such as □, ○, △ or other geometric figures. The selection frame 102 may be blank or have a Corresponding figures or words, such as 1, (2), (A), ©...etc. The selection box 102 can also be set to a positive word distribution, such as a number of strokes to represent its number. Even the selection frame 102 can be set to a matrix of N x M, wherein the N and M are equal to one of 1, 2, 3, 4, 5, and 6, such as a 3x3 dot matrix, for selection Represents the number of meals ordered.

Please refer to the sixth embodiment of FIG. 6 to FIG. 8 , which is a second embodiment of the creation of a quick and easy optical identification order menu editing system, which mainly includes: two customized order menus 100 and 100 ′, and an image recognition module 30 . An image correction module 40, an editing module 50 and the same database 60, wherein the two customized order menus 100 and 100', one for the selection frame 102 to retain the original blank order list 100 The other one of the two order menus 100 and 100' has a corresponding image file 110 and 110'; the image recognition module 30 has a recognition ratio. The positioning mark 101 on the two order menu image files 110 and 110' is set to be different from the content of the two order menu image files 110 and 110'. The frame 102 is selected to quickly identify the relative coordinates of all the selection frames 102; wherein the basic operation of the preferred implementation is as follows: pixel value = (Red + Green + Blue) / 3, and even different Weighted to calculate, such as pixel value = .299 * Red + 0.587 * Green + 0.114 * Blue.

The image correcting module 40 performs deviation correction on the maps of the two order menu image files 110 and 110' to obtain a normalized origin and coordinate axis, and The editing module 50 provides a standardized coordinate and an average pixel value of the selection frame 102 for the restaurant manager to subsequently input the association database for linking the selection frame 102. Corresponding meanings, and; the sample database 60, storing the coordinate box of the identified sample, the normalized pixel value, and the corresponding meaning of the selection frame 102; more specifically, the image in the present invention The capture module 10 can capture images of the two order menus 100 and 100' through a CCD or CIS camera lens and store the image files 110 and 110'. The format of the image files 110 and 110' can be color. Or grayscale.

The sample database 60 can be easily established to link the specific coordinates of the selection frame 102, the average pixel value, and the corresponding selection contents on the blank menu 100, such as table number, number of people, meal name, amount, and Information such as quantity. The simple design of the exclusive menu that meets the needs of the home provides customers with convenient services such as ordering, checkout and follow-up meals.

In summary, the novel utility of the present invention has significant progress compared with the prior art, and at the same time, it has industrial availability, fully complies with the patent requirements, and submits an invention patent application according to law. However, the above is only the preferred embodiment of the present invention, and the scope of the present invention is not limited thereto; therefore, the equivalent changes and modifications made by the scope of the patent application and the contents of the patent specification of the present invention are It should remain within the scope of this invention.

Claims (10)

A quick and easy optical identification point menu editing system mainly comprises: a customized blank point menu, an image processing module, an image recognition module, an image correction module, an editing module and the same database The composition, wherein the customized blank point menu has a corresponding image file; the image processing module converts the image file of the blank order menu from a color format to a gray scale format; the image recognition module Identifying the positioning mark and the selection frame on the image file of the order menu, and storing the average gray scale value and the relative coordinates of the positioning mark and the selection frame in the blank point menu; the image correction module, the image file of the order menu The map is corrected for deviation to obtain the normalized origin and coordinate axes, and the coordinates of the selection frame are standardized; the editing module provides the standardized coordinates of the edited selection frame and the average grayscale difference for the restaurant manager Subsequent input of the association database for linking the corresponding meanings of the selected frame, and the sample database storing the standardized coordinate values and the average gray scale of the selected sampling frame Value, as well as the election draw box corresponding to the meaning; thereby, the food and beverage industry can design their own editing, simple restaurant to complete their own quick and easy optical recognition carte menu. For example, in the fast-track optical identification menu editing system described in claim 1, wherein the image file of the blank order menu is further formed by an image capturing module through a camera lens to capture an image of a blank order menu. . For example, the quick and easy optical identification point menu editing system described in claim 2, wherein the image format of the blank order menu is one of color and gray scale. For example, the fast and easy optical identification point menu editing system described in claim 1 , wherein the editing module further defines a standardized coordinate of each of the selection frames recognized by the image recognition module for input selection. The content includes one of the table number, the number of people, the name of the meal, the amount and the quantity, and forms an associated data structure. For example, the fast and easy optical identification point menu editing system described in claim 1 wherein the sample database stores the standardized coordinates and the average gray scale difference of each of the selection frames, and the corresponding selection includes the table. One of the number, number of people, name of the meal, amount and quantity. The invention relates to a quick and easy optical identification point menu editing system, which mainly comprises: two customized order menus, an image recognition module, an image correction module, an editing module and the same database, wherein the Two customized order menus, one for the original blank order list, and the other for the selected order list, and the two order menus each have a corresponding image file; the image Identifying the module, identifying the content of the two-point meal image file, identifying the positioning mark, and setting a portion with a significant difference in pixel values as a selection frame, thereby generating relative coordinates of the selection frames; The image correction module performs offset correction on the maps of the two order menu image files to obtain a normalized origin and coordinate axis, and normalizes the coordinates of the selection frame; The editing module provides a standardized coordinate and an average pixel value of the editing frame for the restaurant manager to subsequently input the related database to link the corresponding meaning of the selected frame, and the sample database The standardized coordinate value, the average pixel value, and the corresponding meaning of the selection frame of the selected sampling frame are stored; thereby enabling the restaurant industry to design and edit the simple and easy to complete the restaurant's exclusive quick and easy optical identification menu. . For example, in the fast-track optical identification point menu editing system described in claim 6, wherein the image files of the two order menus are further passed through an image capturing module, and the two orderings are taken through the camera lens. A single image is formed. For example, in the fast-track optical identification point menu editing system described in claim 7, wherein the image format of the two order menus is one of color and gray scale. For example, in the fast-track optical identification point menu editing system described in claim 6, wherein the editing module further defines a standardized coordinate of each of the selection frames recognized by the image recognition module for input selection. The content includes one of the table number, the number of people, the name of the meal, the amount and the quantity, and forms an associated data structure. For example, the fast and easy optical identification point menu editing system described in claim 6 wherein the sample database stores a standardized coordinate and an average pixel value of each of the selection frames, and the corresponding selection content includes a table number, One of the number, the name of the meal, the amount and the quantity.