KR20030007230A - The implementation of new layer structure and its control shemes for the processing of images in a single document, and computer-readable media which record the above implementation - Google Patents

Abstract

PURPOSE: A method for executing and controlling a new layer capable of processing an image in a single region through multi-step and a computer-readable recording medium storing a program are provided to offer and realize a new layer and a new layer list structure adaptable to all image processing in order to fit to biometrics developing tool. CONSTITUTION: In case that a layer list storing event is occurred, a name of a storing file is clarified and the file is opened(S2). The information, such as a list ID, a sequential number and a layer number in the list, relating to the layer list of the opened file is written to the file(S3). A position of the first layer in the layer list is searched(S4). In case of an end of the list, the file is closed and the execution is completed(S8,S9). If not, the layer is stored by calling a routine to store the layer in a present position(S6). The position to store the layer at the next time is searched until the end of the list(S7).

Description

A novel layer implementation method and its control method which enables multi-stage processing of images within a single area, and a computer-readable recording medium which is programmed and recorded therein.The implementation of new layer structure and its control shemes for the processing of images in a single document, and computer-readable media which record the above implementation}

The present invention relates to a core technology of a development tool (SIK) for the development of a biometric technology for authenticating / identifying an individual, and more specifically, a layer which is a structure necessary to construct a single platform supporting biometric technology development. The implementation of the layer list, the storage method thereof, and the association control technique between the layers in the list and the program are provided to provide a recording medium.

In general, biometric technology for personal recognition stores a human body's characteristics as a file in a computer through an input device (image input device such as a general camera, a digital camera, a PC camera, a camcorder, etc.), and then uses various image processing techniques. Is a technology that processes and recognizes. As the global market continues to expand, the need for development tools for efficient biometric technology is increasing.

Biometric technology has been evaluated as a breakthrough technology that can solve the problems of security technology so far due to its convenience, accuracy and usefulness. For this reason, there is a growing interest in development tools that can be used to develop biometric security technologies as well as biometric methods worldwide, and are actually being researched and developed by several companies at home and abroad.

Despite these efforts, no effective development tool has been developed so far, since this kind of development tool has fundamentally different characteristics from the existing image processing tool (editing function).

There are two main methods used to develop biometric technology. The first method is a method of individually processing the steps shown in FIG. 1 using different software.

One). Storing characteristics of a living body to be used for recognition through an input device (such as a camera) in a computer ((1) of FIG. 1),

2). Extracting a feature required for recognition of the stored image using an image processing technique ((2) of FIG. 1),

3). Comparing the extracted features with the features of each individual stored in the database, it has a complicated structure of repeating the step (Fig. 1 (3)).

These procedures are commonly applied regardless of the recognition object. Especially, in the case of step 2), finding the optimal characteristics by applying various processing techniques is a key element in the biometric process.

In this case, the results obtained in each step are closely related to each other as having the result of the previous step as their input, and each step should also have information on the image processing technique applied along with the result image.

In addition, this method uses a variety of software to store intermediate results in the form of different files, which makes it difficult to understand the linkage of each step. When developing a system for other biometric information, another software must be used. Because of this, effective development cannot proceed. Moreover, if the file formats supported by each software are different, the conversion process must be performed step by step, and the difficulty is increased because there is a lack of supported image processing techniques.

Therefore, in the case of development tools, it is possible to repeatedly perform complex image processing techniques multiple times in a single area and input images from an input device in one program (S / W). There is a problem that requires image management for multi-stage processing, control, storage, control technology of the processed multi-stage image.

In addition, the existing image editing tools, which are based on simple editing functions, do not provide multi-step complex image processing, and have limitations in adjusting and controlling each result of multi-step iteration processing. In addition, there is a problem in that it is not easy to construct a development tool because there is no essential technology required for a development tool supporting smooth development.

The second method is for system developers to configure and use the software required for their own development. This method has the advantage that all development work can be done with a single software, unlike the first method, but there is a limit to implementing all the required functions.

Accordingly, an object of the present invention is to provide a method for implementing and implementing a new layer and a new layer list structure that can be applied to all image processing so as to be suitable for a biometric development tool.

In addition, the present invention provides a method of storing a new layer and a new layer list structure as a single file that can be applied to all image processing to be suitable for a biometric development tool, and between images in a layer list. The purpose is to provide a method for automatically controlling the association.

In addition, the present invention provides a method for storing a single file of a new layer and a new layer list structure provided to be applicable to all image processing so as to be suitable for a biometric development tool and within a layer list. It is an object of the present invention to provide a computer-readable recording medium by programming a method for automatically controlling the association between images.

On the other hand, the definition of the existing layer is "acetate paper wrapped on one screen", which is a technology that creates a complex image by arranging background → image 1 → image 2 on one screen. It looks like a stack. However, the new layer in the present invention is very different from the existing layer that is superimposed with an "independent structure having all the information on one of several images present on one screen" to make one image. have.

In addition, the layers of the present invention containing the contents of an image have an interconnected structure in the order of generation, which is called a layer list (hereinafter, referred to as a 'list').

1 is a general flowchart for developing a biometric authentication system.

2 is a block diagram of a single platform according to the present invention.

Figure 3 is an example of the internal structure and storage of the integrated storage file in the present invention,

3a). A photo that saves the result of the five steps of the currently read image in a single file.

3b). Photo showing the result saved as eye.dav file.

3c). Photo showing the results of reloading the saved results.

4 is a photograph showing an example of automatic control of the association of the results of the multi-step process in the present invention,

4a). Result photo after 5 steps of read image processing.

4b). Photograph showing that after processing another layer 0, a new layer is created between existing 0's and 1's, and subsequent results are automatically converted.

4c). Photo shows that when you delete an inserted Layer 6, it automatically recalculates to its previous state.

5 is an example of a macro function for image processing developed in the present invention.

5a). Picture which saves image processing information of current list to disk as 'aaa.tsq'.

5b). Photo reading the new image and applying the previously saved 'aaa.tsq'.

5c). Photo showing macro processing result after application.

6 is a data structure diagram of a layer and a layer list developed in the present invention.

7A is a flowchart for storing a multi-stage image processing result in the present invention.

7a). Flow chart for storing multi-level image processing results in a layer list.

7b). Flow chart for storing multi-level image processing results in a layer.

8 is a flowchart illustrating a method for automatically controlling correlation of results processed by multiple steps in the present invention.

9 is a flowchart of a macro function for image processing in the present invention;

9a). The processing flow of 'tsq' file among the image processing macro functions.

9b). Flowchart of macro processing a 'tsq' file in a new image.

Hereinafter, the preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

2 is a configuration diagram of a single platform according to the present invention, 'device independent interface' is a part for controlling a physical input device (camera, optical sensor, PCCAM) used for image acquisition, 'Layer function / file input and output' Is the part that handles the screen output, layer management, and output to file and the image file which is acquired in advance and saved on the disk. 'Signal Processing' is the processing part for the current image. It consists of the processing methods necessary for biometrics such as Fourier transform, Gerber transform, and Harper transform, and 'Decision method' extracts the features necessary for recognition from the transformed image and compares them with existing information in its database. This part is responsible for the actual recognition and consists of simple distance calculation, neural network, and dynamic techniques.

The present invention provides a data structure (layer and layer list) necessary for constructing a development tool supporting biometric system development, and a storage method and a control technique for storing the data structure. A development tool was developed to comprehensively support multi-level processing and recognition.

First of all, due to the development of the layer and list structure, it is possible to easily manage the image required in the situation such as the development of biometric technology, thus enabling the development of a development tool. In addition, since all the contents of the list being processed can be saved in a single file, the development process can be saved and the contents of the list include all the images, parameters, and intermediate results generated during the development process. The development of such storage methods can provide development tools with increased efficiency of technology development. 3 is a diagram showing a process of reloading after saving the processed content to a file as a picture.

Automatic control technology, one of the layers related development technology, solved the image relevance in the development tool. This technique automatically recalculates the content after the inserted or modified step if a new step is inserted or any step is modified in the middle of the result of processing the acquired image in multiple steps. In multi-stage processing of images such as biometrics, the result of the current step serves as the input of the next step. If a new step is inserted, the result of this step must be input again into the existing step to obtain the correct result. The layer automatic control technique of the present invention automatically handles this series of tasks whenever necessary. Figure 4 shows one embodiment for the automatic control of the association of the list (list) developed in the present invention.

And another important technique developed for layers is the macro function for image processing. This technique is described as follows. All the contents of the list created by multilevel processing of the currently loaded image can be stored in a single file as described above. However, if only the image processing information (conversion type, parameters) of the current list can be extracted and stored in a file and applied to a new image collectively, the past processing procedure can be easily reproduced. For example, macro processing similar to those in the 'Hangul 97' program or the 'Hangul 2002' program of the 'Hangul and Computer' company can be applied to image processing. For this function, the present invention newly developed a file format with a 'tsq' extension for storing only image processing information of a list, thereby developing a macro function for image processing.

FIG. 5 shows a process of storing only the conversion result of the current list in a file and applying this information to a newly obtained image by a macro function.

1. Data structure of layers and lists of the present invention.

The structure of the new layer for the multi-stage image processing, which is an essential technology for constructing the development tool, but has not yet been developed, is shown in FIG. 6, and a list thereof is shown in Table 1 below.

Table 1 Implemented Internal Structure of the Inventive Layer List

Structure name Element Field names that make up each element Description of each field Layer Data Structure Control element m_bActive Control variables m_bVisible m_ID Layer id Transform_id Process id obtained from image in layer m_nBitPerPixel The number of bits in the pixels that make up the image m_nChannelCount The number of channels in the image m_nChannelMode Format of the channel m_nEnableBit Control variables m_nWidth; Image horizontal size m_nHeight Image vertical size m_nX Control variables m_nY m_fZX m_fZY m_pLink Connection point for the next layer to be created m_strLayerName The name of the current layer Image storage element m_pData Storage area of actual data (including all areas suitable for all image conversion techniques) m_pDataR m_pDataG m_pDataB m_pGrayData m_pData16 m_pData16R m_pData16G m_pData16B m_pDataRe m_pDataIm m_pDataReR m_pDataReG m_pDataReB m_pDataImR m_pDataImG m_pDataImB m_pData16In m_pBitmap m_pBitmapR m_pBitmapG m_pBitmapB Factor strArgName; Parameter name used in the transformation from which the image was obtained dArgValue; Actual value of the parameter

Table 1 shows a data structure of a layer capable of storing all information about one resultant image. The layer structure in the present invention has been directed to a generalized structure that can be commonly used for various image transformations. To this end, the overall structure is divided into three substructures: control, image storage, and parameter transfer. The controls are the general field (variable) containing the size (horizontal and vertical) of the image pointed to by the current layer, the number of channels and the image storage format, whether it is currently active (m_bactive), whether it is visible on the screen (m_bvisible), and the layer. It consists of the m_nX, m_nY, m_fZX, and m_fZY variables related to the applied process (which may be different in size from the original image after processing). The image storage area stores the actual processed image. Since it is impossible to know in advance which image will be used for which conversion, the real part image (RE part image) can receive all the results of processing all the existing image formats. , Imaginary region image (IM part image), color image, gray image storage area.

The argument passing element contains both the actual name (strArgName) of the translator method applied to the current layer and the actual value (dArgValue;) of the parameter used. These values are actually important variables for later automatic calculation of the layer.

FIG. 6 is a diagram illustrating a structure of a list in which these layers are correlated with each other. FIG. 6 shows a list of layers configured in a memory after three steps of processing a first layer image. The starting image (layer 0) is the first image read and is given the number 0 first. Step 0 is processed step by step to form new layers. If you apply a specific transformation to layer 0, layer 1 is newly created and the result of the transformation is stored in this layer. If you apply specific transformation to layer 1 again, layer 2 is newly created. However, these generated layers do not exist independently of each other, but are interconnected. The reason is that if we remember only the position of layer 0, it should be a structure (list) that can look at all layers. That is, after layer 1 is processed by processing image of layer 0, the link variable (m_plink) of layer 0 is given the start address of newly created layer 1. This process is repeated for each subsequent layer creation. Therefore, as shown in FIG. 6, a linked list is constructed. The value of the link variable of the last generated layer (layer 4 in FIG. 6) becomes 0 so that it is the end of the list.

As shown in FIG. 6 and Table 1, the individual layer case has a data structure that can be generally applied to all image processing.

2. How to save the contents of the current list to a single file.

7 shows a method of collectively storing all the contents of a list having the results of the work up to now (the result of multi-stage image processing) in a single file.

First, a step (S1) of determining whether or not a layer list storage event has occurred, a step of completing execution (step S9) if the layer list storage event does not occur in the determination step (S1), and the determination step When the layer list storage event occurs in step S1, a step of opening and then specifying the name of the storage file (step S2) and a list unique number that is information related to the layer list of the file opened in the step (S2) (List id), number indicating the order of creation (seq id), writing the number of layers (num) in the list (File) step (S3 step), and the position of the first layer of the layer list (number 0) Finding step (step S4), determining whether it is the end of the list (step S5), and if it is determined that the end of the list in the determination step (step S5), closing the file (step S8) and completing the execution (Step S9) and If it is determined in the determination step (S5) that it is not the end of the list, calling a routine for storing the layer of the currently obtained position and storing the layer (step S6), and ending the list in the step (S6) Until then, the next step of finding the location of the layer to be saved (step S7) is to save the layer list as a single file.

In addition, in the process of storing the layer, a step of determining whether a layer storage event has occurred (step S10), a step of returning if a layer storage event has not occurred in the determination step (S10), and (S14); When the layer storage event occurs in the determination step (S10), storing the control information, size, unique number (id) and layer name of the currently found layer in a file (step S11), and in the step (step S11) Writing the necessary image data to a file according to the condition of the control variable (step S12), and storing information related to the conversion used in the step (step S14) to the file (step S13). It returns to a step (step S14).

3. How to automatically control the association of lists

FIG. 8 illustrates a method of automatically recalculating other layers affected by an inserted / modified layer when a new layer is inserted into a list that is currently being processed or when reprocessing one of the existing layers, thereby maintaining correct association. Show the flow for.

That is, determining whether an event has occurred or not whether automatic conversion of the current layer list is required (step S15), and if execution of the event does not occur in the determination step (S15), the step of completing execution (S27), in the case where a transformation-related event occurs in the determination step (S15), creating a new layer and setting a default value (step S16), and the conversion that requested the input image in the step (S16) And storing the result in the image storage area of the generated layer (step S17), determining whether to insert a new layer at the end of the list in step (S17), and If the new layer is inserted at the end of the list in the determination step (S18), the generated layer is inserted at the end of the list (step S26) and the execution is completed (step S27). ), And when the new layer is not inserted at the end of the list in the determination step (S18), the newly created layer position (Y) and the layer position (X) to be placed before the (Y) and the ( Determining a layer position Z to be placed after Y) (step S19), inserting (Y) between (X) and (Z) in the list (step S20) in the step (step S19); In step S20, the position of the next layer to be stored is searched (step S21), and in step S21, the link value of the current layer and (Y) is matched and the current input image value is set to (Y). (S22), and determining whether or not the end of the list in the step (S22) and the end of the list in the determination step (S23) If it is determined that the execution is completed (step S27) and the determination step (S23 step) If it is determined that it is the end of the list, the current input image is processed using the conversion information stored in the current layer and the result is stored in the current layer (step S24), and the current input image is processed previously in the step (step S24). The result of the multi-step processing is to repeat the determination step (step S23) and the storage step (step S24) until the end of the list by matching the link values of the current layer and the current layer. These associations can be automatically controlled (processed) to find the optimal features quickly and accurately.

4. Implement macro function for image processing

FIG. 9 shows a flow of a macro function of extracting only image conversion information of a layer list currently being worked and storing them as a file (tsq extension) and applying them collectively to a new image.

First, in the storing step, determining whether or not an event has occurred whether or not to store processing information (type, parameter) for each layer applied to the current layer list in the form of a "tsq" file (step S26); When the event for the storage does not occur in the determination step (S26) (step S30) and the step of setting the file name to be stored when the event about the storage occurs in the determination step (S26 step) (S27 step) ), Storing the applied image transformation type of the list currently being worked on in the file obtained in the step (step S27) (step S28), and then storing the transformation parameters applied to each layer constituting the list (S29). After the step) is completed, the storage is completed when the execution completion step (S30) is reached.

In the step of macro-processing the contents of the stored 'tsq' file into a new image, step S31 of first reading an image to be newly processed is executed. Next, determining whether to apply the 'tsq' file (step S32), if the event does not occur in the determining step (step S32), completing the execution (step S37), and the determining step ( Reading the stored file (step S33) when the event for macro processing occurs in step S37), obtaining an image conversion type from the file obtained in step S33 (step S34), and then After the step of obtaining the applied conversion parameter (step S35) and the step of generating the layer list by performing the actual conversion using the obtained information (S36), the execution is completed when the execution completion step (S37) is reached.

The present invention described above is not limited to the above-described embodiments, and various substitutions, modifications, and changes can be made without departing from the technical spirit of the present invention, which is common knowledge in the technical field to which the present invention belongs. It is self-evident for those who have.

As described above, the present invention provides a structure of a new layer and a list required to make a development tool suitable for a development environment that needs to repeatedly process an acquired image such as biometrics, a method of storing it as a single file required for this structure, and automatic association. By developing a technology for control, it is possible to develop a convenient and efficient development tool, so that a superior development tool can be constructed, and thus, the ripple effect on the entire industry has a big advantage.

Claims (11)

In constructing a development tool that supports biometric system development, an image layer is created by overlaying it into an independent structure having all information about one of several images present on a screen, and the image layer is generated. A new layer implementation method that allows an image to be processed in multiple stages within a single area by organizing a list of layers in a concatenated structure. In constructing a development tool supporting biometric system development, the starting image (layer 0) is a first read image and is assigned a number 0 first, and the step 0 is processed step by step to form new layers. And generating a new layer 1 by applying a specific transformation to layer 0 and storing the result of the transformation in this layer, and repeating the process of creating a new layer 2 by applying a specific transformation to the first layer. Step and process the image of layer 0 to make layer 1, and then assign the linked address (m_plink) of layer 0 to the start address of newly created layer 1 A new layer implementation that enables multilevel processing of images within a single, contiguous area. The method according to claim 2; A new layer implementation method that enables the image processing in a single area to be processed by extracting only the image conversion information from the layer list currently being worked on and saving it as a file and applying them to a new image collectively. . The data structure of the development tool that supports the biometric system development is composed of the control elements of the fields constituting each element, the image storage element, and the parameter transfer element. A new layer control method that enables multi-stage processing of images within a single region, characterized by integrated storage in a file. The method according to claim 4; The control elements are general fields (variables) containing the size of the image pointed to by the current layer, the number of channels and the image storage format, whether it is currently active (m_bactive), whether it is visible on the screen (m_bvisible), and the processing applied to the layer. It consists of variables related to (m_nX, m_nY, m_fZX, m_fZY), and the image storage element consists of a real part image (RE part image) and an imaginary area image (IM part image) which receives all the results of processing all existing image formats. ), The color image, and the gray image storage area, and the parameter passing element has both the actual name (strArgName) and the actual value (dArgValue;) of the parameter used in the current layer. New layer control method that enables multi-stage processing of images within a single area. Storing a list of layers in a storage means, determining whether a layer list storage event has occurred, completing execution if a layer list storage event has not occurred in the determining step, and layer list in the determining step In the event of a storage event, opening the file after specifying the name of the storage file, a list unique number (List id) related to the layer list of the file opened in the step, and a number (seq id) indicating the generated order. Writing the number of layers (num) in the list to a file, finding the position of the first constituent layer (number 0) of the layer list, determining whether or not it is the end of the list, and If it is determined that the end is completed, close the file and complete execution; if it is determined that the end of the list is not the end of the list, Calling a routine for storing the layer of the obtained position and storing the layer; and in the step of finding the position of the next layer to be stored until the end of the list, the layer list is stored as a single file. New layer control method that enables multi-stage processing of images within a single area. The method according to claim 6; Determining whether a layer storage event has occurred in the storing of the layer, returning if the layer storage event has not occurred in the determining step, and controlling the currently found layer when the layer storage event occurs in the determining step. Storing information, a size, a unique number (id) and a name in a file, writing the necessary image data to a file according to the conditions of the control variable in the step, and converting the information related to the conversion used in the step into a file A new layer implementation that allows for multiple levels of processing of images within a single region, further including writing to and returning to. Automatically calculating a layer list, determining whether an event related to a transformation has occurred, completing execution if an event regarding transformation does not occur in the determining step, and converting at the determining step Creating a new layer and setting a default value when an event related to the event occurs; processing the input image with the requested transformation in the step; and storing the result in the image storage area of the generated layer; Determining whether to insert a new layer at the end of the list; if the new layer is inserted at the end of the list in the determining step, inserting the created layer at the end of the list and completing execution; If you do not insert a layer at the end of the list, (Y) 'and' Layer position (X) to be placed before (Y) 'and' Layer position (Z) to be placed after (Y), wherein (Y) in the step (X) in the list ) (Z20), a step (S21) of finding the position of the layer to be stored next in the step (S20), and a current layer (Y) in the step (S21). Matching the link value of < RTI ID = 0.0 >) and < / RTI > the current input image value with the image stored in (Y), determining whether it is the end of the list in the step, and not determining whether it is the end of the list. If it is determined that the execution is completed; if it is determined that the end of the list is determined in the determining step, processing the current input image using the conversion information stored in the current layer and storing the result in the current layer; Image and wife New layer control that allows images to be processed in multiple stages within a single area that cycles through the decision and storage steps until the results match and the link values of the current layer and the current layer match to the end of the list. Way. Determining whether or not an event has occurred whether or not to save processing information for each layer that has been applied to the current layer list in a file form, and if the event for saving does not occur in the determining step, the execution is completed. And setting a file name to be stored when an event relating to storage occurs in the determining step, storing the image conversion type applied to the current working list in the file obtained in the file name setting step, and then constructing a list. A new layer control method that allows an image to be processed in multiple stages within a single area consisting of the steps of completing the saving after completing the step of saving the conversion parameters applied to each of the layers. Reading a new image to be processed, determining whether to apply one of the 'tsq' files, completing execution if no event occurs in the determining step, and processing the macro in the determining step Reading the stored file, obtaining the image conversion type from the file obtained in the above step, obtaining a conversion parameter applied to each of the layers, and performing a real conversion with the obtained information. A new layer control method for processing an image within a single area, characterized in that the macro processing the contents of the stored file into a new image by completing the step of generating the list and completing the execution. A computer-readable recording medium in which a new layer implementation and control method for programming an image in a single area according to any one of claims 4 to 10 is programmed.