JP2009134562A - Image processor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To suppress deterioration of a whole image formed by partical images even if swipe speed varies. <P>SOLUTION: It is determined whether read slice image data is input from a line sensor or not (step 104). When it is affirmative, a difference between pixel values of pixels in the same positions of input read slice image data and comparison image data stored in a comparison image data storage area is obtained for whole pixels, and the total number of pixels whose difference is other than zero is obtained (step 106). It is determined whether a value of the total pixels is not more than a threshold or not (step 108). When it is negative, read slice image data is stored in a slice image data storage area to be subjected to stitching (step 110). Comparison image data stored in the comparison image data storage area is deleted. Read slice image data is stored as new comparison image data (step 112). <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an image processing apparatus.

  In a system that performs fingerprint authentication, either an area (surface) sensor or a line sensor is used to acquire a fingerprint image.

  When the area sensor is used, the entire fingerprint image can be obtained by reading once.

  On the other hand, the line sensor continuously captures a plurality of partial images (hereinafter referred to as slice images) in which a fingerprint is divided into strips by sliding (swiping) a finger on the sensor. For this reason, when a line sensor is used, processing for constructing an entire fingerprint image from a plurality of slice images (hereinafter referred to as stitch processing) is performed.

  When the number of pixels in the swipe direction of the slice image is one, the whole image of the fingerprint is constructed by connecting the slice images obtained continuously.

  In Patent Document 1, when a whole image of a fingerprint is constructed by connecting slice images each having one pixel in the swipe direction, the slice image acquired by the previous reading and the slice image acquired by the current reading are described. When the similarity between the two exceeds a predetermined threshold, the slice image acquired by the current reading is retained, and when the similarity does not exceed the threshold, the slice image acquired by the current reading is A technique is disclosed that suppresses the occurrence of redundant portions in the entire fingerprint image by discarding.

On the other hand, when the number of pixels in the swipe direction of the slice image is plural, the entire fingerprint image is constructed by detecting and overlapping the overlapping portions from the slice images obtained continuously.
JP 2003-256816 A

  By the way, when superimposing, a shift | offset | difference may arise in the relative position of slice images. Such a displacement of the relative position is caused by performing a superimposition process in units of pixels, and therefore, a displacement of the overlap between slice images is generated by a maximum of half a pixel.

  In addition, since the whole image of the fingerprint is constructed by further superimposing the slice image on the image constructed by superposition, the generated relative position shift also affects the subsequent superimposition processing.

  The displacement of the relative position causes deterioration of the image quality of the entire fingerprint image such as collapse of the image and extension.

  Further, when a line sensor capable of high-speed scanning is used, when the speed of swipe relative to the scanning speed of the line sensor is relatively slow, slice images with many overlapping portions are necessarily continuous, and The number increases. In such a case, the deviation of the accumulated relative position becomes large. The effect of such accumulation of relative position deviation is particularly problematic when trying to increase the swipe speed adaptation range in a line sensor capable of high-speed scanning.

  The present invention has been made in view of such circumstances, and when a whole image is constructed from a partial image read using a line sensor capable of high-speed scanning, even if the swipe speed varies. An object of the present invention is to provide an image processing apparatus that suppresses deterioration of an image to be constructed.

  In order to achieve the above object, an image processing apparatus according to the present invention includes a reading unit that reads a part of a reading target at every predetermined time interval and generates partial image data, and a storage unit that stores the partial image data. And determining whether the partial image data before registration in the storage unit generated by the reading unit matches the latest partial image data of the partial image data stored in the storage unit And means for registering the partial image data before registration in the storage means as the latest partial image data when the matching degree is lower than a predetermined degree.

  According to the present invention, the reading unit reads a part of the reading target at every predetermined time interval and generates partial image data.

  The determination unit determines the degree of matching between the partial image data before registration in the storage unit generated by the reading unit and the latest partial image data of the partial image data stored in the storage unit.

  The registration means registers the partial image data before registration as the latest partial image data in the storage means for storing the partial image data when the matching degree is lower than the predetermined degree.

  The image processing apparatus according to the present invention may further include a construction unit that constructs an entire image to be read from the partial image stored in the storage unit.

  The determination means may determine the degree to which the entire partial image data before registration matches the entire latest partial image data.

  Further, the determination unit may determine the degree to which the data of at least a part of the pixel area of the partial image data before registration matches the data of the pixel area of the latest partial image data.

  Further, the determination means may determine the degree to which the data of at least a part of the pixel area of the partial image data before registration matches the data of the neighboring area including the pixel area of the latest partial image data. Good.

  According to the present invention, the partial image data is registered only when the degree of matching between the partial image data before registration and the latest registered partial image data is lower than a predetermined degree. It is possible to reduce the number of partial image data used when constructing an entire image to be read from each. Thereby, it is possible to suppress a shift in relative position accumulated in the entire image.

  As described above, the present invention can reduce the number of partial image data used when constructing an entire image to be read from each of the partial images, and suppress a shift in relative position accumulated in the entire image. And has an excellent effect that the deterioration of the entire image to be constructed can be suppressed.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
(First embodiment)
As shown in FIG. 1, a fingerprint reading apparatus 10 according to this embodiment includes a bar-shaped line sensor 12. FIG. 1B is a cross-sectional view taken along the line AA in FIG. The fingerprint reading device 10 reads the entire fingerprint area of the finger by sliding the finger on the line sensor 12 in the swipe direction indicated by the arrow in FIG. .

  The line sensor 12 reads a series of partial images representing a fingerprint from an authentication target finger swiped on the sensor surface. FIG. 1B is a cross-sectional view taken along the line AA in FIG. The line sensor 12 is a sensor surface in which the length in the swipe direction indicated by the arrow in FIG. 1B is shorter than the length of the entire fingerprint of the finger and the width in the direction perpendicular to the swipe direction is approximately the same as that of the finger. have. The line sensor 12 swipes in the swipe direction on the sensor surface, and continuously reads an image (swipe image) of a partial region of the fingerprint region from a finger that moves relative to the sensor surface.

  The fingerprint reader 10 is configured as a part of a mechanism for performing fingerprint authentication processing, for example, and is used by being mounted on a personal computer or the like.

  Next, the configuration of the fingerprint reader 10 will be described with reference to the block diagram shown in FIG.

  In addition to the line sensor 12, the fingerprint reader 10 includes a flash memory 14, an image processing mechanism 24 in which a CPU 16, a ROM 18, a RAM 20, and the like are connected by a bus 22, a mechanism connected to the fingerprint reader 10, and data An interface 26 for performing transmission / reception is included.

  The flash memory 14 stores various data including a template used for fingerprint authentication processing.

  The image processing mechanism 24 performs a construction process for constructing the entire fingerprint image data from a series of slice image data obtained by reading by the line sensor 12.

  An outline of this construction process will be described with reference to FIG.

  When the read slice image data read by the line sensor 12 is input, the image processing mechanism 24 performs a thinning process TT. In this thinning-out process TT, the comparison image data stored in the comparison image data storage area MC provided in the RAM 20 is compared with the read slice image data to determine whether or not to be the target of the stitch process TS. This is a process of thinning out read slice image data that is not the target of the stitch process TS.

  The read slice image data to be stitched by the thinning process TT is used as comparison image data to be used in the next thinning process TT, and is provided in the RAM 20 for storing the slice image data to be stitched TS. In addition, it is stored in the slice image data storage area MS.

  The comparison image data storage area MC has a size that can store all the pixels of one slice of image data. For example, when the slice image data for one sheet is image data in which the horizontal direction is x pixels and the vertical direction is y pixels, and one pixel is represented by s [bits], the size of the comparison image data storage area MC is x Xy * s [bits]. Y is an integer of 2 or more.

  The image processing mechanism 24 performs stitch processing TS, which will be described later, using slice image data stored in the slice image data storage area MS. The whole image data of the fingerprint constructed by the stitch process TS is output to a mechanism for performing the fingerprint authentication process connected to the fingerprint reader 10 via the interface 26 so as to be used for the fingerprint authentication process.

  Next, a processing routine for constructing the entire fingerprint image executed by the image processing mechanism 24 will be described with reference to the flowchart of FIG. A program for this processing routine is stored in the ROM 18 of the image processing mechanism 24.

  In step 100, the comparison image data stored in the comparison image data storage area MC is deleted.

  In step 102, it is determined whether or not reading by the line sensor 12 has been completed, that is, whether or not the finger swipe operation has been completed. If the determination is affirmative, the process proceeds to step 114, and if the determination is negative, the process proceeds to step 104.

  In step 104, it is determined whether or not read slice image data is input from the line sensor 12. If the determination is affirmative, the process proceeds to step 106, and if the determination is negative, the process returns to step 102.

  In step 106, comparison processing between the comparison image data stored in the comparison image data storage area MC and the read slice image data input in step 104 is performed.

  In the comparison processing, the difference between the pixel values of the pixels at the same position between the input read slice image data and the comparison image data stored in the comparison image data storage area MC is obtained for all the pixels, and the total number of pixels where the difference is not 0 Find DiffNum.

  In step 108, it is determined whether or not the value of the total number DiffNum obtained in step 106 is equal to or less than a threshold value THDiffNum. If the determination is affirmative, the process proceeds to step 102, and if the determination is negative, the process proceeds to step 110.

  In addition, what is necessary is just to determine the value of threshold value THDiffNum based on an experimental result.

  That is, when the value of the total number DiffNum is equal to or less than the threshold value THDiffNum, the image area represented in the read slice image data is the image area represented in the slice image data that is already the target of the stitch process TS. The slice image data is substantially the same and is determined to be the slice image data to be thinned out, and the read slice image data is not stored as the target of the stitch process TS.

  On the other hand, when the total number DiffNum exceeds the threshold value THDiffNum, the image area represented in the read slice image data is different from the image area represented in the slice image data that is already the target of the stitch process TS. It is determined that the region is an object of stitch processing TS.

  That is, in step 110, the read slice image data is stored in the slice image data storage area MS as a target of stitch processing TS.

  In step 112, the comparison image data stored in the comparison image data storage area MC is deleted, and the read slice image data is stored as new comparison image data.

  In step 114, stitch processing TS is performed.

  The outline of the stitch process TS will be described with reference to FIG.

  First, an overlapping position between an image already constructed by superposition and a slice image is obtained. As shown in FIG. 5A, the slice image PS is relatively moved in units of pixels with respect to the image PA that has already been constructed by superposition, and the degree of correlation of the overlapping regions of each image is determined for each movement. The position where the degree of correlation is maximized is calculated.

  Next, the image already constructed by the superposition and the slice image are superposed. As shown in FIG. 5B, for the pixels in the area A1 and the area A2 where the image already constructed by superimposition and the slice image do not overlap, the pixel values of the image already constructed by superposition Alternatively, the pixel value of the slice image is adopted. On the other hand, for the pixels in the region A3 where the image already constructed by superposition and the slice image overlap, the pixel value is the average value of the pixel value of the image already constructed by superposition and the pixel value of the slice image. And

  By repeating the above process for the number of slice images stored in the slice image data storage area MS, an entire image of one fingerprint is constructed.

  As described above, in the present embodiment, when the area of the image represented in the read slice image data is substantially the same as the area of the image represented in the slice image data that is already the target of the stitch process TS, By not using the read slice image data as the target of the stitch process TS, it is possible to suppress unnecessary superimposition processing.

  Also, by reducing the number of slice image data to be processed by stitch processing TS, it is possible to suppress the deviation of the accumulated relative position, and to suppress the deterioration of the overall image quality of the fingerprint, such as the occurrence of low contrast dullness. Is possible.

Furthermore, since thin slice image data having a large degree of overlap is selected and thinned out instead of simply thinning out, it is possible to perform appropriate thinning that matches the speed of finger movement.
(Second Embodiment)
Next, the construction process according to the second embodiment of the present invention will be described.

  In the first embodiment, the case has been described in which the thinning process TT is performed by comparing the entire slice image data, and the slice image data that is the target of the stitch process TS is selected. In the second embodiment, A case where image data of a part of slice image data is compared will be described. In the following, description of portions corresponding to those of the first embodiment will be omitted, and differences from the first embodiment will be mainly described.

  In the second embodiment, the size of the comparison image data storage area MC provided in the RAM 20 is set to a size that can store pixels of slice image data for one line in the slice image data.

  For example, when the slice image data is image data in which the horizontal direction is x pixels and the vertical direction is y pixels, and one pixel is represented by s [bits], the size of the comparison image data storage area MC is x × s [bits. ]. Y is an integer of 2 or more.

  Next, a processing routine for constructing an entire fingerprint image executed by the image processing mechanism 24 according to the second embodiment will be described with reference to the flowchart of FIG.

  In step 150, the comparison image data stored in the comparison image data storage area MC is deleted.

  In step 152, it is determined whether or not reading by the line sensor 12 has been completed. If the determination is affirmative, the process proceeds to step 164. If the determination is negative, the process proceeds to step 154.

  In step 154, it is determined whether or not read slice image data is input from the line sensor 12. If the determination is affirmative, the process proceeds to step 156. If the determination is negative, the process returns to step 152.

  In step 156, comparison processing between the comparison image data stored in the comparison image data storage area MC and the pixel row data of the Nth line of the read slice image data input in step 154 is performed.

  N may be determined by an arbitrary method.

  In the comparison processing, the difference between the pixel values of the pixels at the same position as the comparison image data stored in the data comparison image data storage area MC of the pixel row of the Nth line with the input read slice image data is equivalent to one line. And the total number DiffNum [N] of pixels having a difference other than 0 is obtained.

  In step 158, it is determined whether or not the value of the total number DiffNum [N] obtained in step 156 is equal to or less than a threshold value THDiffNum. If the determination is affirmative, the process proceeds to step 152. If the determination is negative, the process proceeds to step 160.

  In addition, what is necessary is just to determine the value of threshold value THDiffNum based on an experimental result.

  That is, when the value of the total number DiffNum [N] is equal to or smaller than the threshold value THDiffNum, the image area represented in the read slice image data is the image of the image represented in the slice image data that is already the target of the stitch process TS. The area is substantially the same, and it is determined that it is slice image data to be thinned out, and the read slice image data is not stored as a target for stitch processing TS.

  On the other hand, when the total number DiffNum [N] exceeds the threshold value THDiffNum, the image area represented by the read slice image data is the image area represented by the slice image data that is already the target of the stitch process TS. It is determined that the region is different from that of the stitch processing TS.

  That is, in step 160, the read slice image data is stored in the slice image data storage area MS as the target of the stitch process TS.

  In step 162, the comparison image data stored in the comparison image data storage area MC is deleted, and the pixel row data of the Nth line of the read slice image data is stored as new comparison image data.

  In step 164, the stitch process TS described above with reference to FIG. 5 is performed.

Thus, in the present embodiment, the amount of memory used as the comparison image data storage area MC in the thinning-out process TT is suppressed by storing the pixel row data of the Nth line of the slice image data and obtaining the difference. It is possible to reduce the amount of calculation processing for obtaining the difference.
(Third embodiment)
Next, the construction process according to the third embodiment of the present invention will be described.

  In the second embodiment, the case of comparing the data of the pixel row for the Nth one line of the slice image data has been described, but in the third embodiment, the image of the L line of the slice image data is used. A case of comparing data will be described. In the following, description of portions corresponding to the second embodiment will be omitted, and differences from the second embodiment will be mainly described.

  With reference to the flowchart of FIG. 7, a processing routine for constructing the entire fingerprint image executed by the image processing mechanism 24 according to the third embodiment will be described.

  In step 200, the comparison image data stored in the comparison image data storage area MC is deleted.

  In step 202, it is determined whether or not the reading by the line sensor 12 has been completed. If the determination is affirmative, the process proceeds to step 220. If the determination is negative, the process proceeds to step 204.

  In step 204, it is determined whether or not read slice image data is input from the line sensor 12. If the determination is affirmative, the process proceeds to step 206. If the determination is negative, the process returns to step 202.

  In step 206, a constant “N” is substituted for the line counter “i”.

  In step 208, it is determined whether or not the line counter “i” is equal to or smaller than a constant “N + L−1”. If the determination is affirmative, the process proceeds to step 210. If the determination is negative, the process returns to step 202.

  Note that N and L may be determined by an arbitrary method.

  In step 210, the comparison image data stored in the comparison image data storage area MC is compared with the pixel row data of the i-th line of the read slice image data input in step 204.

  In the comparison process, the difference of the pixel value of the pixel at the same position as the comparison image data stored in the data comparison image data storage area MC of the pixel row of the i-th line with the input read slice image data is equivalent to one line. And the total number DiffNum [i] of pixels having a difference other than 0 is obtained.

  In step 212, it is determined whether or not the value of the total number DiffNum [i] obtained in step 210 is equal to or less than a threshold value THDiffNum. If the determination is affirmative, the process proceeds to step 214, and if the determination is negative, the process proceeds to step 216.

  In addition, what is necessary is just to determine the value of threshold value THDiffNum based on an experimental result.

  In step 214, the line counter “i” is incremented by “1”.

  In step 216, the read slice image data is stored in the slice image data storage area MS as a target of stitch processing TS.

  That is, when any value of the total number DiffNum [i (N ≦ i ≦ N + L−1)] exceeds the threshold value THDiffNum, the image region represented by the read slice image data is already the target of the stitch process TS. It is determined that the region is different from the region of the image represented in the slice image data and is the target of the stitch process TS.

  When all the values of the total number DiffNum [i (N ≦ i ≦ N + L−1)] are equal to or smaller than the threshold value THDiffNum, the image region represented by the read slice image data is already the target of the stitch processing TS. The area of the image represented by the slice image data is substantially the same and is determined to be the slice image data to be thinned out, and the read slice image data is not stored as the target of the stitch process TS.

  In step 218, the comparison image data stored in the comparison image data storage area MC is deleted, and the pixel row data of the Nth line of the read slice image data is stored as new comparison image data.

  In step 220, the stitch process TS described above with reference to FIG. 5 is performed.

  As described above, in the present embodiment, the pixel row data of the Nth line of the read slice image data is stored, and the difference for the maximum L lines is obtained, so that the comparison image data storage area MC is used in the thinning process TT. The amount of memory to be used can be reduced, and only slice image data generated by reading different areas exceeding the L line is subjected to the stitch process TS. Therefore, the averaging process performed when composing the entire fingerprint image is performed. It is possible to suppress the image quality degradation caused by it.

  In each of the above-described embodiments, the configuration of the entire image of the fingerprint has been described as an example. Is applicable.

(A) is the schematic of the fingerprint reader which concerns on embodiment of this invention, (B) is AA sectional drawing of (A). 1 is a block diagram of a fingerprint reading apparatus according to an embodiment of the present invention. It is a figure which shows the outline | summary of a construction process. 6 is a flowchart of a processing routine for constructing an entire image of a fingerprint according to the first embodiment. (A) is a conceptual diagram of the process for obtaining the overlay position, and (B) is a conceptual diagram of the overlay process. It is a flowchart of the process routine of construction of the whole image of a fingerprint concerning a 2nd embodiment. It is a flowchart of the process routine of construction of the whole image of the fingerprint concerning a 3rd embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Fingerprint reader 12 Line sensor 14 Flash memory 24 Image processing mechanism

Claims (5)

Reading means for reading a part of the reading object at predetermined time intervals and generating partial image data;
Storage means for storing the partial image data;
A determination unit that determines the degree to which the partial image data before registration in the storage unit generated by the reading unit matches the latest partial image data of the partial image data stored in the storage unit; ,
A registration unit that registers the partial image data before registration in the storage unit as the latest partial image data when the matching degree is lower than a predetermined degree;
An image processing apparatus.   The image processing apparatus according to claim 1, further comprising a construction unit that constructs an entire image to be read from the partial image stored in the storage unit.   3. The image processing apparatus according to claim 1, wherein the determination unit determines a degree of matching between the entire partial image data before registration and the entire latest partial image data. 4.   The said determination means determines the degree to which the data of the pixel area of the at least one part of the partial image data before the registration and the data of the pixel area of the latest partial image data match. Image processing apparatus.   The determination unit determines a degree of matching between data of at least a part of pixel area of the partial image data before registration and data of a neighboring area including the pixel area of the latest partial image data. The image processing apparatus according to claim 1 or 2.

Images