JPH0685179B2 - Image processing program generation method - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for generating an image processing program, and more particularly to a method for generating an image processing program with improved efficiency in image processing.

[Background of the Invention]

2. Description of the Related Art Image processing apparatuses in recent years have advanced at an accelerating pace, with a focus on improving processing speed and expanding processing functions. In such an image processing apparatus, it is desired to improve the processing speed, but in terms of expanding the processing functions, it is necessary to train an olter capable of selecting the functions in a short time. So, in order to enhance the man-machine interface, a conversational processing method is essential,
A conventional conversational processing method is a method of inputting a command, a command and a parameter, or displaying a command on a menu screen and selecting it by a number.
However, in such a conventional method, it is necessary to know in advance the relationship between the command and the processing result of the command, and it is the fact that a person without prior knowledge cannot actually use this sufficiently. Furthermore, in this method, the processing result can be obtained only after inputting and executing the command, so it is necessary to set again even if there is a mistake in input or a slight difference in parameter setting. It took a considerable amount of time even for a skilled person to find out.

The one described in Japanese Patent Application Laid-Open No. 55-6640 is known as the one adopting such a system.

[Object of the Invention]

An object of the present invention is to provide an image processing program generation method capable of easily creating an image processing program without requiring any skill because the parameters can be adjusted interactively without inputting image processing commands one by one. Especially.

[Outline of Invention]

In the present invention for achieving the above object, first, an image is stored in advance as a test pattern image (hereinafter, referred to as a fixed image pattern), and the fixed image pattern is processed and displayed in real time in a predetermined order. (Catalog display) If the target processed image, the commands and parameters are automatically stored, and by repeating this procedure, a series of processing procedures can be created easily and in a short time. Is. That is, the plurality of image processing commands are divided into a plurality of image processing command groups including a plurality of related image processing commands, and the plurality of image processing commands in each image processing command group are processed in advance in descending order of frequency of use. The order is set, a menu screen including a plurality of the image processing command groups is displayed on the output device, one image processing command group is selected from the menu screen by a conversation type, and the selected image processing is performed. Image processing is performed on the test pattern image according to the first image processing command in the command group, and each time a search command is input in an interactive manner, the image processing command is changed in accordance with the processing order to change the test pattern. Image processing is performed on the image, and the processed image is the same as the original image of the test pattern image. It is displayed in contrast to one screen, and while watching the displayed processed image, the parameters of the corresponding image processing command are finely adjusted in a conversational manner, and when the desired processed image is obtained, it is used for the image processing. The image processing command and the adjusted parameter are registered in the command registration table as a registered image processing command.

Example of Invention

Hereinafter, examples of the present invention will be described, but before that, a technique on which the present invention is based will be described.

FIG. 1 is a block diagram showing a technology as a basis of the present invention and a device configuration of an embodiment of the present invention. The image recognition / analysis device includes an input device 1, an image processing device 2, an output device 3, an image processing / saving file device 4, and an interactive input device 5.

Image processing in such a device is roughly divided into area,
In order to obtain the feature amount such as the perimeter, it is divided into a pre-process of previously performing binarization, thinning, noise removal, etc. on the original image and a post-process of obtaining the feature amount. Since there are various types of commands for image processing, use the interactive image evaluation system (SHIN / IE) to interactively extract commands for pre-processing and post-processing and set the accompanying parameters. There is.

As a concrete example of this, as shown in FIG. 2, one command is set at a time (step 100), parameters are set (step 101), and the command is executed (step 10).
2) The image processing is executed and evaluated each time (step 10).
3) Determine whether or not all processing is completed (step 104) and register (step 105) the key-in mode, and display the menu screen as in the processing flow shown in FIG. 3 (step 110). Select the command (step 11
1) Set parameters (step 112), then select next command (step 113), execute creation command (step 114), and judge the processing result (step 11).
5) There is a mode of registering the result (step 116), that is, a menu mode of interactively selecting commands and executing image processing from the menu screen. And in each mode, or this 2
Image processing was performed by combining the two. Such image processing will be described with reference to FIG.

First, in order to obtain a desired processing result, it is necessary to obtain an image processing command that matches the purpose and further determine its parameters. Therefore, how to obtain this command (step 120 in FIG. 4) is shown in FIG. 5 (I).
Will be described with reference to.

First, the contents of the processing (image capture, color image processing, grayscale image processing, etc.) are determined on the menu screen shown in FIG. 5 (I) (C _{N at} level L ₁ ). Then, select a command that the user thinks the best according to their content, i.e., at level L ₂ C _NM, a C _NM1 Level L _3.
Until this command is selected, there are usually three levels (levels L ₁ , L ₂ ,
L ₃ ) step is needed. After deciding a command, enter P _MNK at level L ₄ and P _NMK2 at level L ₅ on the parameter setting screen as a parameter peculiar to the command, and finally enter the parameter P for one command C _NM1 .
_You can decide up to _NMK2 . Image processing can be performed on the original image _only after the parameter P _NMK2 is determined. (Level L6). The command thus obtained is created in step 120, the command is temporarily registered (step 121), and image processing determination is performed (step 123). If this step 123 is different from the expected one, the parameter input and execution processing on the parameter setting screen must be repeated again (steps 120 to 1).
twenty three). In the worst case, it may be necessary to reset the command and start again. Since about 10 commands are required to extract the feature amount, the user needs to repeat the parameter input from the menu screen through the parameter setting screen and the series of about 10 image processes.

This will be explained based on an actual example.

FIG. 6 shows how the contour line of the flaw 7 of the book piece 6 is taken out. FIG. 7 shows a processing flow of the operation of extracting the contour line. When the book piece 6 shown in FIG. 6 (a) is taken in through a television camera (step 140) and an image is displayed on the monitor (step 141), FIG. 6 (b) is displayed.
Is displayed as. When the contour extraction processing is performed in step 143, a screen with the contour emphasized is obtained as shown in FIG. 6 (c). Further, in step 144, the noise on the screen is removed and binarized to reduce it as a normal screen (see FIG. 6 (d)). When the screens of FIGS. 6 (c) and 6 (d) are calculated between binary images (step 145) and the contour of the flaw is extracted (step 146), the screen shown in FIG. 6 (e) is obtained.

If this is processed using the interactive image evaluation system which is the basis of the present invention, it becomes as shown in FIG. First, select the processing contents on the menu screen of step 150 to enter the command selection and setting mode (step 151),
The command "SDIS" is set from the command group 9.
After this, this parameter is set (step 152).
This is the code 11 in the parameter group for the command group 9.
As described above, "03" and "01" are input, registration and image processing are performed (step 153), and the original image is displayed on the output device for the first time. According to the determination result, the processes of steps 151 to 154 are performed, and it is further determined whether or not all the processes are completed (step 155), and if completed, it is registered (step 156).
In this example, command selection is required 10 times as a whole, so that parameter input, registration, and image processing related to this need to be repeated at least 10 times or more.

As described above, the technique which is the basis of the present invention requires considerable prior knowledge for command selection and parameter setting, and various parameter settings can be considered.
It is difficult for a person skilled in the art to set uniquely. For this reason, it is necessary to repeat the resetting and re-inputting of the parameters, and the target image intended by the user cannot be easily obtained. Even if the time required for actual computer processing is shortened by hardware and the speed is increased, it cannot be said that the functions of the high-speed image processing apparatus cannot be fully utilized by repeating the input / output as described above. It was

The biggest network to achieve this speedup is
The time required for command selection and parameter setting is long, that is, the turnaround time for the entire image processing depends on the level of each user.

The cause of this may be that the processing result is displayed on the output device after the input setting of the parameter, the execution of the registration / image processing are completed, and the determination must be made for the first time based on this image.

As described above, since the technology that is the basis of the present invention has a configuration that is too command-aware, in addition to having to follow the procedure of “command determination → parameter determination → image processing execution”, Meaning, it was necessary to have enough parameter content as prior knowledge,
Further, it is inconvenient that it takes a considerable amount of time to create an algorithm, such as the necessity of inputting parameters every time.

FIG. 1 is a block diagram showing an embodiment of the present invention, which has already been described and will be omitted. FIG. 9 shows an embodiment of the present invention in comparison with its underlying technology. FIG. 9 (I) is an example of the basic technology, and FIG. 9 (II) is this embodiment. is there. As can be understood from the figure, according to the present embodiment, the man-machine function 13 is arranged to enter the image processing device 14 side.

By the way, in the basic technique, as shown in FIG. 5, the parameter setting is performed by selecting the (image processing) command on the menu screen, and then the target command is obtained by visually confirming the processed image. As already mentioned, a wealth of image processing knowledge is required to select each command and set its parameters, and the turn-around time of interactive image recognition largely depends on the level of the user. It was.

Therefore, according to the embodiment of the present invention, after the grouping number is selected on the menu screen as shown in FIG. 5 (II), the human element (reference numeral 13 in FIG. 9 (II)) is maximally eliminated. The screen corresponding to the selection is immediately displayed, and the image corresponding to the intended image is finely adjusted so that the command corresponding to each image is automatically created. That is, in this embodiment, it is not necessary to be aware of the image processing command and its parameters.

FIG. 10 shows a processing configuration of the present invention.

That is, in the processing configuration, one image of the test pattern image is prepared, the test pattern image is processed by a preset command group, and the fixed pattern processing (blocks 200 to 203) for displaying and the actual processing target image ( For the ITV input image or the image previously stored in the storage device),
The command registration table 204 for storing the commands and parameters automatically generated corresponding to each processing result image, which is divided into the target pattern processing (blocks 200, 206, 207, 208) that performs the same processing as the fixed pattern processing, and the contents of the command registration table It is composed of the device 205 for saving and recovering the data.

Next, a detailed description will be given with reference to FIGS. 10 to 13.

In FIGS. 10 and 11, first, a processing pattern image is selected (step 210) and (block 200). Next, the grouping number is set by the menu screen display (step 211). The fixed pattern processing and the target pattern processing have the same processing procedure except that the processing targets are different. Therefore, the fixed pattern processing will be described here. In addition,
FIG. 11 shows the processing flow of FIG. 10, FIG. 12 shows a list of commands that the operator can use in the embodiment, and FIG. 13 shows a specific example of visual confirmation,
The figure (I) shows a processing example of a fixed pattern, and the figure (II) shows a processing example of a target pattern.

If you specify a fixed pattern (11th Is a return key) and a menu screen is displayed to simultaneously display the representative processing contents of the grouped image processing commands corresponding to the grouping numbers 1 to n (step 21).
1). A structure is used for this grouping, and the subordinate structure of each group (image processing command group) has a sequential structure.

Next, the grouping number which seems to be the desired processing is set (step 213, block 202). In the relevant group, empirically, the processing order is set to the least frequently used image processing command and the least frequently used image processing command. The image processing command at the head of the group corresponding to the set number is automatically processed and the processing result image is displayed (step 213, block 202). As for the display method, as shown in FIG. 13 (I), consideration is given so that the right half of the fixed pattern image can be processed and compared with the original image. When the displayed processing result image is different from the target image, the next processing result image is sequentially displayed by the sequential search command. When the target image is obtained, fine adjustment processing is executed as necessary (step 214, block 203). Here, the fine adjustment is to obtain the optimum processed image by changing the value of the parameter added to the image processing command. The parameter before the fine adjustment is a value most frequently used as a default value, or an average value is set in a case such as a threshold (threshold) when obtaining a binary image. On the processing result display screen, the image processing command name and the number of parameters are displayed together, and fine adjustment can be performed sequentially from parameter No1.
The parameter No is specified by using numeric keys (parameter specification command). The parameter change uses both the balance judgment tree and the sequential method.

First of all, the change from the default value to the intermediate value in the ascending direction of the parameter value is "+" by using the method of the balance judgment tree.
You can specify the change to the intermediate value in the descending direction with the "-" key. Furthermore, when changing from that point one by one, it can be realized by designating the "F" key in the ascending direction and the "B" key in the descending direction. When the fine adjustment is completed, use the temporary registration command or the registration command to enter the command registration table 20.
Can be registered in 4. Command registration table 204
Is configured to automatically count case numbers each time it is registered.

By repeating the above operation, the work of the program becomes easy without inputting a command.

On the other hand, for the target pattern, a program created with a fixed pattern is actually evaluated using raw image data from an industrial camera or the like, or the fixed pattern is not used from the beginning, and the program is directly processed by the same processing procedure as the fixed pattern. Can be created. The program created with the fixed pattern is automatically stored in the buffer and then executed to display the processed image. Therefore, if further fine adjustment is desired, it can be performed by the same method as the fixed pattern. When the desired image is obtained, register it. By repeating this, the program can be easily evaluated.

In addition, by providing functions for list output display of created programs, sorting, deletion, and execution of arbitrary commands in the list as necessary, comprehensive editing, execution,
Evaluation is possible.

Further, by providing a function of saving (SAVE) and restoring (LOAD) the created program in the external storage device, it is possible to expand the program resources.
Also, by adding a file name to the program, management of program resources becomes easy.

As described above, by utilizing the advantages of the high-speed image processing apparatus and displaying the processing result images of various commands in a catalog in real time, it is possible to realize a conversational high-speed image processing apparatus that can easily create a processing procedure without being aware of the command. .

It is also possible to use a voice input / output device or the like instead of the input / output device using the keyboard described in this example.

FIG. 14 and FIG. 15 show the present embodiment and its underlying technology in comparison, and according to the present embodiment (B _B .T _A ), as can be understood from the drawings, It would be considerably shortening treatment with basic technical procedure B _B and its convergence time T _B.

〔The invention's effect〕

As described above, according to the present invention, it is possible to easily create an image processing program without requiring skill because the parameters can be adjusted interactively without inputting image processing commands one by one.

[Brief description of drawings]

1 is a block diagram showing an embodiment of the present invention, FIG. 2 is an explanatory diagram shown for explaining an example of a key-in mode, FIG. 3 is an explanatory diagram shown for explaining an example of a menu mode, and FIG. FIG. 7 is a diagram showing an outline of a technical process which is the basis of the present invention, FIG. 5 is a supplementary explanatory diagram of the command creation of FIG. 4, FIG. 6 is an explanatory diagram of a processing procedure according to an embodiment, and FIG. FIG. 8 is a flow chart of the processing procedure, FIG. 8 is an operation flow chart of the same basic technique, FIG. 9 is a diagram for explaining the points of interest of the embodiment of the present invention, and FIG. 10 is a process schematic diagram of the same embodiment. FIG. 11 is a processing flow chart of the same embodiment, FIG. 12 is an explanatory view showing management commands, FIG. 13 is an explanatory view showing an image display example, and FIG.
FIG. 15 is an explanatory diagram shown for comparing the system according to the basic technology with the system of the present embodiment, and FIG. 15 is a characteristic diagram showing an outline of the time required for the present invention and the basic technology. 1 ... Input device, 2 ... Image processing device, 3 ... Output device, 4 ... Image data storage device, 5 ... Conversation evaluation type image recognition device input device, 13 ... Man-machine function, 14 ... Image Processor, 15 ... Management command list.

 ─────────────────────────────────────────────────── ─── Continuation of front page (72) Inventor Toshio Usui 5-2-1 Omika-cho, Hitachi-shi, Ibaraki Inside the Omika factory, Hitachi, Ltd. (56) Reference JP-A 61-112285 (JP, A) JP-A-60-179875 (JP, A)

Claims (1)

[Claims] 1. A method for generating an image processing program, which captures image data to be processed, and performs image processing according to a plurality of set image processing commands and parameters respectively associated with the image processing commands. It is divided into a plurality of image processing command groups consisting of a plurality of related image processing commands, and the plurality of image processing commands in each image processing command group are set in advance in descending order of frequency of use. A menu screen including the image processing command group of is displayed on the output device, one image processing command group is selected from the menu screen by a conversational type, and the first image in the selected image processing command group is displayed. Performs image processing on the test pattern image according to the processing command, and searches interactively Each time a command is input, the image processing command is changed according to the processing order to perform image processing on the test pattern image, and the processed image and the original image of the test pattern image are displayed on the same screen. When the desired processed image is obtained by finely adjusting the parameters of the corresponding image processing command while watching the displayed processed image, the image processing command used for the image processing is displayed. A method for generating an image processing program, characterized in that the adjusted parameter is registered as a registered image processing command in a command registration table.