JPH07284014A - Picture handling device - Google Patents

Abstract

PURPOSE:To highly accurately detect high movement by selecting the area of high contrast as a representative point area on object pictures at the time of fetching contrast information relating to the prescribed block area of picture data and evaluating it by using a function part. CONSTITUTION:In a camera, for the picture data fetched to a detection circuit 51, that are luminance signals, prescribed high frequency components are extracted by using a band pass filter 51a, a processing is performed in an absolute value processing circuit 51b and then, absolute value processing output is added in a cumulative adder 51c. Contrast data are written in addresses corresponding to respective division areas as contrast values in a memory 51d, the data of the memory 51d are used by a CPU 51e and the position information of the division area R75 of the high contrast is outputted. In such a manner, the division area R75 is detected in a contrast detection circuit 51 and an area positioned at a center in the area is set as a feature area Bf.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image handling device, and more particularly, to an image handling device having a function unit for taking in and evaluating contrast information regarding a predetermined block area of image data.

[0002]

2. Description of the Related Art Conventionally, as an image handling apparatus having a function part for taking in and evaluating contrast information regarding a predetermined block area of image data, one subject image is divided by scanning a movable mirror, for example. A camera is conceivable in which the divided subject image is captured by an image pickup element and the image data of the divided subject image is pasted together at the overlapping portion to generate image data for one subject image. Further, this type of camera can capture high-definition captured image data corresponding to an effectively obtained 1-frame (field) image even if an image sensor having a normal resolution is applied. It is something.

In this camera, the above-mentioned divisional image data is required to be overlapped as close as possible to each other, but it is possible that the combined portion will be slightly misaligned between photographing due to mechanical scanning accuracy. Therefore, it is necessary to detect the deviation amount for each photographing and perform the bonding based on the deviation amount.

Therefore, a possible bonding process starts from the block area P on the first bonding portion R11 of the first image-taking screen M11 (see FIG. 3), which is a reference for bonding, as shown in FIG. The two first and second characteristic regions Ba and Bb to be obtained are located on the second bonding section R12 of the second photographing screen M12 (see FIG. 3) to be bonded, and Of the two corresponding areas Bc and Bd which have a high correlation and are to be overlapped with each other, and the deviation amount thereof is detected. Then, by moving the second screen relative to the first screen so that the characteristic regions Ba and Bb and the corresponding regions Bc and Bd of the second screen are overlapped, the bonding unit The image data of the photographing screen M10 (see FIG. 3) for one subject image such that the image is inconspicuous is generated. The overlapping of the above-mentioned areas is performed by the characteristic area Ba,
Characteristic region points Ba0 and Bb0 which are predetermined upper left points of Bb and corresponding region points Bc0 which are predetermined upper left points of corresponding regions Bc and Bd,
This is done by matching Bd0.

The image data stitching technique is not limited to the split photographing type camera for stitching the images, and is also applied as a technique for detecting a shift amount between two subject images, that is, a motion vector amount. It is a thing.

By the way, FIG. 51 shows a reference image and a target image in a motion detection process of a two-image data screen in a video camera which is a conventional image handling apparatus, and an enlarged area view of representative points (feature points). is there.

In this conventional example, a representative point (feature point) matching method is applied, the reference image M71 is an image serving as a matching reference for motion detection, and the target image M72 is an image after movement. And The reference image M71 and the target image M72 are divided into four regions. Then, the representative point area B72 set on each of the four divided screens of the target image M72 is composed of, for example, 8 × 8 pixels. The search area R71 in which the image data corresponding to the representative point area B72 should be searched is the reference image M.
71 is set on the 4-split screen.

In this conventional example, the image data of the image is binary data, and the sum of the image data 0 and 1 corresponding to the black and white portions of the representative point area B72, that is,
A block area composed of 8 × 8 pixels equal to the cumulative addition value Σ is detected as a corresponding point from the search area R71 of the reference screen M71, and the memory space address of the detected corresponding point area and the representative point area B72. The amount of movement due to the movement, that is, the amount of deviation is calculated from the memory space address of.

[0009]

However, in the case of the above-mentioned conventional example, the representative point area (characteristic area) B72 is set at a fixed position regardless of the contrast state of the pattern of the target image, such as a white area. , A black portion may be set, and does not necessarily have image information suitable for search. Therefore, there is a problem that the accuracy of motion detection does not increase.

The present invention has been made to solve the above-mentioned problems, and it is possible to quickly select a high-contrast area as a representative point area on a target image, and it is possible to perform highly accurate motion detection. It is an object of the present invention to provide an image handling device including

[0011]

SUMMARY OF THE INVENTION According to the present invention, an image related to the one area and an image related to the other area that partially overlap each other are overlapped in at least one area in relation to the overlapping area. To perform the bonding process so that the area set as one feature part in the area to be registered normally overlaps with the area recognized as one corresponding part in the overlapping area related to the other areas that should correspond to these areas. An image handling apparatus having a functional part, wherein when setting a characteristic part, an area to be set in itself is divided into a plurality of blocks larger than the characteristic part, and the characteristic part is set in the block. The first discriminating operation as to whether or not the predetermined first requirement such as the contrast value of the image corresponding to the setting is performed is executed as a relatively rough area discriminating operation in each block unit, and then the first discriminating operation is performed. Discriminating action of The second predetermined requirement such as the contrast value of the image corresponding to the setting of the feature in the area within the block only for the block discriminated as satisfying the predetermined requirement corresponding to the setting of the feature in the self The image handling device is provided with a setting means for executing a second discriminating operation for identifying a small region satisfying the above condition as a relatively dense region discriminating operation for the intra-block region.

[0012]

According to the present invention, when the setting means sets the characteristic part, the area in which the characteristic part is to be set is divided into a plurality of blocks larger than the characteristic part, and the characteristic part is set in the block. The first discriminating operation as to whether or not the predetermined first requirement corresponding to is satisfied is executed as a relatively coarse area discriminating operation in each block unit, and thereafter, the first discriminating operation is performed in the self by the first discriminating operation. A second discriminating operation for specifying a small area satisfying a predetermined second requirement corresponding to setting as a characteristic portion for an area in the block only for a block discriminated to satisfy a predetermined requirement corresponding to setting a section Is executed as a relatively dense area discriminating operation for the area in the block.

[0013]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a main block configuration diagram of a camera showing a first embodiment of the present invention.

The camera according to the present embodiment is a camera which is an image handling device suitable for recording characters, characters, photographs, etc., written on a blackboard or the like, and a high-definition captured image. It is a split-screen camera that can be obtained. The captured image data forming one frame (field) is formed by dividing the image capturing area of the subject in the horizontal direction of the screen by scanning the movable mirror, and the divided areas are composed of two CCDs arranged vertically. The image pickup optical system 2 having a two-plate configuration is used to take in each divided area. And FIG.
As shown in (1), the overlapping position of the captured images of the divided screen areas M1 to M4 of the areas A to D divided in the horizontal direction is combined to correct the scanning position deviation for each shooting, and one shooting screen ( (1 frame) M0 image data.

In the divided screens M1 to M4, overlapping bonding areas R1 to R shared by adjacent screens are shared.
Have six. Further, the divided photographed image data of the divided screens M1 to M4 is composed of photographing outputs of two systems, respectively, and it is necessary to perform joining processing of the images of the upper and lower two screens M1a and M1b. Since there is no deviation, it is not necessary to correct each time of shooting. Further, the captured image based on the main video signal obtained by the above-mentioned bonding process is displayed on the monitor, but is further recorded as image data on the recording medium. A memory card, a floppy disk, a hard disk, a magneto-optical disk, or the like can be applied as this recording medium. It can also be output to a printer or the like and printed out.

As shown in FIG. 1, in the present camera, a subject image is scanned by a mirror scanning section 1 composed of a rotating mirror mechanism, and a divided image is formed by an image pickup optical system 2 having a two-plate structure for each divided subject area. Two-system image pickup signals for each divided screen M1 to M4 (see FIG. 2) in each area are output to the monochrome process circuit 3. The output is captured in the frame buffer 4 as image data of a split screen. The frame buffer 4 also performs mirror inversion processing, non-interlace processing, and misalignment correction processing at the time of bonding performed via the first memory controller 16.

Interpolation processing is performed by the interpolation calculation circuit 5, and the output is processed through the gamma correction circuit 6 and the pseudo halftone correction circuit 7. Then, the image correction output is
It is input to the changeover switch 10 controlled by the image area separation circuit 15. The image data output from the change-over switch 10 is compressed by the compression circuit 11 controlled by the CPU 12 and written in the memory card 22 via the card 1 / F 13.

The output of the interpolation calculation circuit 5 is a binarization circuit 9 whose threshold is set by the edge emphasis circuit 8 and the CPU 21.
Image data of two pasting portions, which are the overlapping portions of the two divided screens to be pasted in the first half binarized image signal, which are binarized by the second memory controller 19 With the write address being controlled by, the data is written in the boundary buffer 17 which is a memory for calculating the amount of deviation in the bonding described later.

Further, the output of the interpolation calculation circuit 5 is input to an AF (automatic focusing), AE (automatic exposure) processing circuit 14 and an image area separation circuit 15. AF, AE processing circuit 14
Is output to the CPU 21 and automatic focusing or automatic exposure control is executed. Further, the switching operation of the selector switch 10 is controlled by the output of the image area separation circuit 15. That is, the frame buffer 4
When it is determined that the interpolated output from is the natural image data, the output of the pseudo halftone correction circuit 7 is fetched, and when it is determined that it is the character image data, the output of the binarization circuit 9 is fetched. Controlled.

The correlation calculation processing circuit 18a is based on the image data taken in the boundary buffer 17 and the image data of the two bonding portions which are bonded to each other.
At 2, the correlation calculation as the motion vector for the two image data is performed, and the shift amount for bonding is calculated by the shift calculation circuit 18b. The shift amount is based on the previous screen (which is the first sheet in the first bonding process),
The subsequent (second sheet in the first bonding) screen shift is given by the movement amount (horizontal movement amount x, vertical movement amount y), the rotation amount (θ), and the zoom ratio. Details will be described later with reference to the block diagram of FIG. The movement amount (horizontal movement amount x, vertical movement amount y), the rotation amount (θ), and the zoom ratio indicating the above-mentioned shift amount are C
It is input to the first memory controller 16 controlled by the PU 21, and the write address of the image data in the frame buffer 4 is corrected. The first memory controller 16 also performs mirror distortion correction and non-interlace processing in the frame buffer 4 in addition to the shift amount correction.

Next, regarding the correlation calculation and the shift calculation processing,
This will be described with reference to FIG. 3, which is a diagram showing a bonding screen, and FIG. 4 which is a block diagram around the detection circuit 18 including the correlation calculation processing circuit 18a and the deviation calculation circuit 18b.

FIG. 3A shows two divided screens M11 and M12 to be pasted together, and FIG. 3B shows a photographing screen M10 which is a screen after pasting. In addition, in the present embodiment, the subject image is divided into four and photographed in FIG. However, in the following description, the process of pasting the two screens together will be described in detail, but the same process is repeated for pasting more screens.

In the state of FIG.
In the state after the captured image data of the first image is captured, the boundary buffer 17 stores 2 of the pasted portion R11 of the captured screen M11 before the reference (described as the first image).
The binarized image data of the pasting portion R12 which is the overlapping portion of the subsequent (explained as the second image) photographing screen M12 to be pasted together with the binarized image data is fetched.

By the way, in the conventional split screen pasting camera, the characteristic regions Ba and Bb for obtaining the correlation with the corresponding regions are all white (for example, region Be in FIG. 37) or all black. In the case of a screen of a region having no contrast, it is unsatisfactory to search the corresponding region. Therefore, in order to set the characteristic regions Ba and Bb on the photographing screen, it is necessary to check the contrast state of the corresponding region.

As a checking method, as shown in an enlarged view of the screen area in FIG. 38, a block composed of a block pixel number nB (5 × 5 pixels in this case) having a predetermined number of pixels to be detected. The pixel data Ds of each pixel in the area P is 0, 1 data, in this example, a monochrome image is used, and black is detected as 0 and white is detected as 1 in the data obtained at a predetermined threshold. Then, a cumulative addition value ΣP (DS) that is the sum of the values is obtained as an index indicating the degree of contrast. In this case, it is assumed that the block area P is formed by 5 × 5 pixels, and the added value Σp (Ds)
Is expressed by the following equation. That is,

[0027]

## EQU1 ## Σp (Ds) = Dk + Dk + 1 ... ...... + Dk + 24. Then, the regions P having the predetermined number of pixels nB, which is as close as possible to 1/2 of the 25 pixels in this case, are selected as the characteristic regions Ba and Bb.

Therefore, the cumulative sum value Σp (Ds) of the total sums for each block area is defined as one row of pixel positions in the upper portion or the lower portion of the pasted portion of the first photographing screen. Alternatively, it is changed column by column, and as shown in FIG. 39, Σ1 (Ds) ... Is obtained from Σ0 (Ds). The Σ0 (Ds) and Σ1 (Ds) are shown by the following equations. That is,

[0029]

[Equation 2] Σ0 (Ds) = D0 + D1 ………… + D24

[0030]

[Equation 3] Σ1 (Ds) = D5 + D6 ... + D29.

In this example, Σm (D
s) is calculated by the above (Equation 1), and Σp (Ds)
Is searched for a high-contrast region having a high degree of contrast, which is close to 1/2 of the number of pixels nB, and is set as the first characteristic region Ba or the second characteristic region Bb.

FIG. 40 is a cumulative addition arithmetic circuit diagram for performing the above cumulative addition. A region P (p) of pixel m columns × pixel n rows of the image memory 511 designated by the control unit 510.
(= 0) pixel data is cumulatively added by a cumulative adder 512 to obtain a cumulative addition value Σp (Ds). When the cumulative calculation of the region is completed, the reset signal RST is reset, and the calculation for the regions P (p = 1), P (p = 2) ... Is sequentially executed.

FIG. 41 shows a time chart of the image data output signal and the RST signal of the memory 511 and the output signal of the cumulative addition value Σp (Ds).

However, in the conventional cumulative addition calculation processing by the camera, the calculation of the above (Equation 1) is performed every time the region p is changed by one row or one column, and the processing time becomes extremely long. Therefore, a method is conceivable in which the cumulative addition value Σp (Ds) is obtained by cumulative calculation using the cumulative addition calculation circuit of FIG. 42 and the first and second characteristic regions are set.

This calculation method uses the cumulative addition value Σp (Ds)
Is calculated by the following formula. That is,

[0036]

Σp (Ds) = Σp −1 (Ds) − (Dk + ... + Dk +
4) + (Dk + 25 + ... + Dk + 29) However, in this calculation, only the value Σ0 (Ds) in the area P (p = 0) including the first row is (Equation 1), and therefore the addition operation is performed by (Equation 2). The following (Formula 4) is followed for the region P (p = 1) and the subsequent regions which are lowered by one line. That is,

[0037]

[Formula 5] Σ1 (Ds) = Σ0 (Ds)-(D0 + ... + D
4) + (D25 + ... + D29) The above (Equation 5) is Σ 0 (D
s), the sum of the pixel data of the first row (D0 + ...
+ D4) is subtracted and the sum value of new additions (D25 + ... + D29) is added to obtain the cumulative addition value Σ1 (Ds). The same applies to the following areas. In this way, each block area is searched and the first feature area or
The second characteristic region is set.

FIG. 42 is a cumulative addition arithmetic circuit diagram for performing the above cumulative addition. The control unit 5 is different from that of FIG.
This is a circuit in which a multiplier 513 with a changeover switch, which can change ± 1 by 10 SEL signals, is added. Other configurations are the same as those in FIG. This circuit carries out the calculation of (Equation 4).

FIG. 43 shows a memory 51 based on the circuit of FIG.
The time chart of the image data output signal of No. 1, the SEL signal, and the output signal of the cumulative addition value Σp (Ds) is shown.

However, even in this arithmetic processing,
Every time the detection block area p is changed by one line, (D
Addition of k + ... + Dk + 4) and (Dk + 25 + ... + Dk + 2)
Since it is necessary to add 9), the calculation time becomes very long. Then, when the number of pixels of the unit area when selecting the first characteristic area or the second characteristic area is increased, the calculation time further extends, and the first characteristic area of the high contrast area in a short time. Or, the second characteristic region cannot be selected, and as a result, it takes too much time to obtain the image data forming one frame (field).

Therefore, in the present embodiment, a calculation for evaluating the degree of contrast of an image in the block area by sequentially moving the block area to be evaluated can be executed very quickly, and high-definition image data can be handled quickly. Is possible.

Next, in the correlation detection circuit 18a which constitutes the detection circuit 18, as described in the pasting process in the conventional camera with reference to FIGS. 37 to 43, the pasting of the first photographing screen M11 is performed. Part R11 has m rows ×
A value for setting the first characteristic region Ba and the second characteristic region Bb having the size of a block region of n rows, in this case, 5 × 5 pixels, and evaluating the degree of contrast of these characteristic regions, That is, it is a block region showing an equal value by the cumulative addition value Σp (Ds) obtained by adding the pixel data shown in (Equation 4). Laminated part R1 of the second shooting screen M12
2 on the second corresponding area Bc or the second corresponding area B
Search for d.

Then, in the shift calculating circuit 18b constituting the detecting circuit 18, the characteristic region point Ba0, which is a predetermined upper left point on the first characteristic region Ba and the second characteristic region Bb,
Bb0, the first corresponding area Bc and the second corresponding area Bd
A displacement amount at the time of pasting for matching the corresponding area points Bc0 and Bd0 which are the upper left predetermined points is calculated. This shift amount is indicated by a movement amount (horizontal movement amount x, vertical movement amount y), a rotation amount (θ), and a zoom ratio.

The camera of this embodiment is characterized by the method of setting the positions of the first characteristic region Ba and the second characteristic region Bb, but the details of the processing will be described later.

The shift amount data is output to the first memory controller 16 and the second image data written in the frame buffer 4 is corrected based on the shift amount. After that, the bonding section for the second image data is written on the R11 side of the boundary buffer 17, and the bonding section for the third image data is written on the R12 side, and the operation is repeated. Similarly, the amount of deviation is detected for the divided photographing screens that are taken in thereafter, and the pasting is performed for each.

FIG. 5 is a diagram showing a block configuration of a portion for outputting the cumulative addition value Σp (Ds) in the correlation detection circuit 18a of the detection circuit 18 of FIG.

In the camera of this embodiment, the first,
When setting the second characteristic regions Ba and Bb, the calculation is simplified by using the register to obtain the cumulative addition value Σp (Ds) regarding the reference detection block region P, the calculation time is shortened, and the quick calculation is performed quickly. 1, second characteristic region B
a and Bb can be set.

That is, the detection block area P is the same m columns × n rows of pixels as the block area shown in FIG. 38, in this case 5 × 5 pixels, and cumulative addition is performed by adding each pixel data 0, 1. Find the value and move it by a rows. In the present example, the above-mentioned cumulative addition value is obtained by moving each row in the vertical direction. Then, the calculated value Σ0 (Ds) as shown in (Equation 2), (Equation 3), (Equation 4), (Equation 5), etc.
), Σ1 (Ds), ... Σp (Ds) is calculated. At this time, in the case of the present embodiment, the block area 1 shown in FIG.
The value of (Dk + ... + Dk + 4), which is the partial sum per line,
Registers (1) 18a2 and registers (2) corresponding to rows sequentially
18a3, ... Write to register (n) 18a4 and add those values by adder 18a5 to obtain Σp (Ds
). Since n = 5 in this embodiment, the register (n) will be referred to as the register (5) in the following description.

More specifically, (Equation 1) is rewritten to obtain (Equation 6). That is,

[0050]

[Equation 6] Σp (Ds) = Dk + Dk + 1 + 1 …… + Dk
+ 24 = (Dk + ... + Dk + 4) + (Dk + 5 + ... + Dk
+9) + (Dk + 10 + ... + Dk + 14) + (Dk + 15 + ...
+ Dk + 19) + (Dk + 20 + ... + Dk + 24) Write the partial addition value of each row in () on the right side shown in register (1) to register (5) respectively, and add the values of those registers by the adder 18a5. The cumulative addition value Σ
Output p (Ds).

FIG. 6 shows the cumulative addition value Σp (D
s) shows the time chart for calculating. An enlarged waveform of memory output and the like is shown in the lower part of the figure. The image data of the target block area is shown in FIG.

When the cumulative addition is started, the memory 17 is initially set.
To output pixel data D0 to D24 for the number of pixels in the block area m × n, and store the added output at the Q point in the registers (1) to (5) for each m data of one row. At the time of register switching, the second memory controller 19
A CNT signal is output, and the register numbers 1 to n, that is, 1 to 5 are switched in order to write the added value of each row. Also, each time, RS from the second memory controller 19
It is assumed that the addition output of the adder 18a1 is reset by the T signal. Then, the value stored in each register is added to the adder circuit 18a5.
The cumulative addition value Σ0 (D
s) is calculated.

Therefore, the second memory controller 19
Output the CNT signal that rewrites the register data,
The register number of the storage destination of the partial sum data is switched from n, that is, from 5 to 1. Therefore, when obtaining the cumulative addition value Σ 1 (Ds) in the next block shifted by one row, the pixel data of one new row, for example, the sum value of D25 to D29 is stored in the register (1). It Then, the values stored in the registers are added by the adder circuit 18a5 to obtain the cumulative addition value Σ1.
(Ds) is output. Below, shift the area line by line,
The cumulative addition value Σp (Ds) is sequentially obtained while forming the overlapping portion of the block areas.

According to the present embodiment, when the cumulative addition value Σp (Ds) of the block area designated by sequentially moving within the predetermined range is calculated, when the designated position is moved by one line at the overlapping portion of the block areas, , Rewriting the register value,
The cumulative addition value Σp (Ds) can be obtained by a small number of processes in which the sum of pixel data for one row is stored in the register and the value of the register for the number of rows in the block area is added. Therefore, the above-mentioned cumulative addition value Σp (Ds) quickly detects a block area exhibiting high contrast, and the area can be selected as the first and second characteristic areas Ba and Bb, and the bonding with higher accuracy can be performed. Can be executed.

FIG. 8 shows the number of calculation processes for obtaining the cumulative addition value Σp (Ds) in the camera of the present embodiment, which is 2
It shows a comparison with the case of the camera described in the example of the technology of the conventional example of the kind. In the figure, m and n are columns of the block area,
The number of rows is shown, and k is the number of search block areas for which the cumulative addition value Σp (Ds) is obtained. As shown in the figure, in the case of the present embodiment, the number of processing times is a fraction.

In the above embodiment, the data extraction unit for obtaining the cumulative addition value Σp (Ds) is one pixel which is a unit of the block area, but the cumulative addition value Σp is several pixels as one unit. (Ds) may be obtained.

Further, the unit of movement of the block area for calculation does not have to be limited to one column and one row.
With respect to the block area of n rows, a natural number a smaller than m or n may be moved in the column or row direction by an amount obtained by multiplying the dimension of the unit area (pixel) in the corresponding direction.

Furthermore, the moving direction is not limited to the horizontal direction or the vertical direction, and any direction such as an oblique direction is included in the concept of the present invention.

Next, the bonding process in this embodiment,
That is, the first and the first images set on the first screen as a reference.
Match the second characteristic regions Ba and Bb with the first and second corresponding regions searched on the second sheet screen to be bonded, and
A process of pasting the first and second split screens together will be described.

FIG. 9 shows two divided screens M11 and M12.
FIG. 11 is a diagram showing a screen M10 obtained by pasting them together.
This bonding is performed by setting the set first and second characteristic regions B
When a and Bb are overlapped with the first and second corresponding areas Bc and Bd searched on the second sheet, as shown in the figure, characteristic points Ba0 and Bb0 on the first and second characteristic areas Ba and Bb , And the corresponding points Bc0 and Bd0 on the first and second corresponding areas Bc and Bd are made to coincide with each other and are bonded.

FIG. 10 is a diagram showing a bonding state when the first and second corresponding points Bc0 and Bd0 are deviated. 1
When the actual corresponding points Bc0 and Bd0 of the second screen M14 are deviated from the corresponding points Ba0 ′ and Bb0 ′ of the second screen M13 which do not have positional deviations corresponding to the characteristic points Ba0 and Bb0, Stick together as shown.

FIG. 11 is a view showing a pasting processing state of the above-mentioned screens M13 and M14. As shown in (A), first, the first feature point Ba0 and the first corresponding point Bc0 are set. Shift to match. Then, as shown in (B), the second screen M14 is rotated to match the second feature point Bb0 with the second corresponding point Bd0 and perform the bonding.

The image handling apparatus of this embodiment can quickly select an area having high contrast as the representative point area on the target image, and enables highly accurate motion detection.

FIG. 24 is a block diagram of the motion detecting section of this image handling apparatus. As for the image data output from the imaging system, first, the captured image data of the first field is written in the field memory 52 as reference image data. At the same time, the information about the contrast value in the image data of the first field is detected by the contrast detection circuit 51 together with the area information and taken into the motion detection circuit 53.

After that, the picked-up image data of the second field is taken in as the target image data in the motion detection circuit 53, is read from the field memory 52 at the same time, and the reference image data of the first field is taken in in the motion detection circuit 53. .

Therefore, the high contrast portion is detected from the contrast information for the image data of the first field, which is taken in by the motion detecting circuit 53.
The detection of the high-contrast portion is, for example, to detect the divided area R75 in the divided screen of the image data screen M75 of FIG. And
High contrast division area R75 Position information as it is 2
It is applied as high-contrast position information for the first image data.

This assumes that there is no large shift in the high contrast area between the first and second images, and the contrast information of the first image is used for the second image.

Then, as shown in FIG. 27, the divided area R25 showing the high contrast for the second target image data.
Further, the region at the center position of is designated as the characteristic region Bf. Further, in the motion detection circuit 53, in the image data of the first reference screen M74, the accumulation indicating the high contrast state of the characteristic region Bf from the image data of the area R76 corresponding to the high contrast area R75. A corresponding block area in which the added value Σ matches is selected as a corresponding area. The amount of movement of the second sheet with respect to the first sheet is calculated from the amount of relative displacement between the corresponding area and the characteristic area Bf, and is output as motion vector information. This motion vector information is stored in the first memory controller 1 shown in FIG.
6, and the image data in the frame buffer 4 is rewritten.

FIG. 25 shows a time chart of the contrast information fetching process of the image data of the motion detecting section of FIG. As the image data, when the target image data is sequentially output as A field data, B field data, ..., Contrast data in each field is taken into the motion detection circuit 53 via the contrast detection circuit 51. At the same time, the image data of each field is also taken into the field memory 52.

Immediately after reading the image data, during the vertical synchronizing signal output period, the high contrast position information of the target image is read from the captured contrast data. In the next field period, the CPU outputs the position designation information on the target image.

FIG. 26 is a diagram showing a detailed block configuration of the contrast detection circuit 51 of the motion detection section of FIG. 24 and signal waveforms during each processing. A predetermined high frequency component is extracted from the image data (luminance signal) taken into the detection circuit 51 by a BPF (band pass filter) 51a, processed by an absolute value processing circuit 51b, and then an absolute value processing output is made by a cumulative adder 51c. 27 and the contrast value is added.
The address is written in the contrast data memory 51d at the address corresponding to each divided area. Further, the CPU 51e outputs the position information of the high contrast divided region R75 (see FIG. 27) from the data of the contrast data memory 51d.

FIG. 27 is a diagram showing a division state of the captured image data screen. In the apparatus of this embodiment, the contrast detection circuit 51 detects the division area R75 having the high contrast. . Then, the region located in the center of the area is set as the characteristic region Bf.

As described above, according to the image handling apparatus of the present embodiment, when setting the characteristic region in the target image, first, the contrast data for each divided area of the target screen is taken in, and the high contrast among them is acquired. Since the divided area indicating is detected and the central area of the divided area is set as the characteristic area, it is possible to specify an ideal high-contrast area for the characteristic area used at the time of detecting the corresponding area later. It becomes possible to detect.

As a modified example of the above-mentioned image handling apparatus with respect to the method of setting the characteristic region described with reference to FIG. 27, the method shown in FIG. 28 is applied to the high contrast division area R75 detected by the contrast detection circuit 51 of FIG. Further, it is a method of surely searching for an area showing contrast in the area. That is, the area R75 is divided into a predetermined number, for example, 12 as shown in FIG. The division area P
By sequentially moving c, the contrast information of each area is detected, and the area showing high contrast is set as the characteristic region Bg.

According to this method, it is possible to surely search the area showing high contrast in the divided area R75.

As another modification of the characteristic region setting method, the method shown in FIG. 29 is used in the case where a high contrast region is searched for in the high contrast divided area R75 detected by the contrast detection circuit 51 of FIG. For example, it is assumed that the divided area Pfm consists of columns × n rows of pixels, and the contrast information of each area is detected while sequentially moving the row or the column by a pixels, and the area showing high contrast is defined as the characteristic region. It is set as Bg.

According to this method, it is possible to surely and finely search the area showing high contrast in the divided area R75.

Next, another example of the image handling apparatus for the motion detection technique related to the above-mentioned division of the divided images will be described.

FIG. 30 is a block diagram of the motion detecting section of the proposed image handling apparatus. The configuration of the image pickup device system and the like other than that shown in FIG. 30 is the same as the configuration of the camera of FIG.

The image data output from the image pickup system is processed by the binarization circuit 61, and the binarized image data of the overlapping portion (bonding portion) Rh of the reference first picked-up image data, Binary image data of the overlapping portion (bonding portion) Ri of the second captured image data to be superimposed is written in the boundary buffer 62 as target image data and reference image data.

In the contrast detection circuit 63, a predetermined size in the first area Rh, for example, 100 × 100.
The contrast information of the image data (0, 1) for each block area of the pixel is detected, and the two areas having high contrast are loaded into the motion detection circuit 64 with the positional information of the areas Rh, Ri exhibiting high contrast. Further, a region narrower than the size of the regions Ph and Ri, for example, in a block region of 8 × 8 pixels, a region showing high contrast due to pixel data (0, 1) is detected, and the region is divided into two characteristic regions B.
Set as h and Bi.

Further, within the regions Rj and Rk in the second image corresponding to the regions Ph and Ri, the block information corresponding to the contrast information (cumulative addition value of image data) of the characteristic regions Bh and Bi is searched. Then, two corresponding areas Bj and Bk are obtained. Then, the amount of movement and the amount of rotation when the corresponding regions Bj and Bk of the second screen are overlapped on the characteristic regions Bh and Bi of the first screen are output as displacement amounts. The shift amount is input to the first memory controller 16 shown in FIG. Then, the data of the second screen captured in the frame buffer 4 is moved and rotated based on the shift amount to perform superposition.

FIG. 31 shows the contrast detection circuit 6 described above.
3 is a block configuration diagram showing the processing states of the detection circuit of the high contrast region in the motion detection circuit 64. The processing state of this circuit will be described as an operation for detecting contrast data for a block area of 8 × 8 pixels, for example.

Now, considering the contrast data of the image data of the two areas of the search block areas Pi and Pj, it is assumed that the image data is the data shown in (A) and (B) of FIG. First, the cumulative adder 65 of FIG. 31 causes the image data (0,
1) is accumulated and the values 60 and 30 are obtained. In the subtractor 66,
The value 32, which is 1/2 of the number 64 of pixels in the block area, is subtracted from the output values 60 and 30 to obtain the values 28 and -2. The absolute value conversion circuit 67 calculates the absolute value, and the subtracter 68 subtracts the absolute value 28,2 from the value 32 which is ½ of the number of pixels in the block area to determine the contrast height. 4, 30 are output as the contrast data shown. Based on this output value, as the contrast state of the search block regions Pi and Pj, it is determined that the region Pj has higher contrast and the region Pi has lower contrast.

The contrast data of the block area Pk shown in FIG. 32C shows that each pixel is almost black (0) and the contrast is low.

Next, still another image handling device having a motion detecting technique related to pasting of divided images will be described.

FIG. 33 is a block diagram of the motion detecting section of the proposed image handling apparatus. This detection unit has substantially the same configuration as that shown in FIG. The difference is that the contrast detection circuit 51A is used for AF (automatic focusing) control of the photographing optical system other than the process of detecting the contrast information of the block area of the target image data. The AF section of the AF / AE circuit 14 of FIG. 1 corresponds to the contrast detection circuit 51A.

FIG. 34 shows a division state of the AF area M81 on the photographing screen for AF processing, and one division block is 30 ×.
The block area R81 is composed of 96 pixels, and the contrast information of a predetermined area is used for AF control. Further, in the case of this modification, this divided area is used as it is for detecting the characteristic region of the image superimposing processing.

The contrast detection circuit 5 shown in FIG.
1A has substantially the same configuration as the contrast detection circuit 51 of FIG. 26, and as described above, the CPU 51e also outputs contrast information for AF.

According to this modification, since the contrast detection circuit 51A is used for both AF control and detection of the characteristic region of the image superimposing processing, the circuit scale does not increase.

Next, still another image handling device using the motion detection technique related to the division of the divided images will be described.

FIG. 36 is a block diagram of the motion detecting section of the proposed image handling apparatus. This detection unit has substantially the same configuration as that shown in FIG. The difference is that the contrast detection circuit 63A is also used for AF (automatic focusing) control of the photographing optical system, other than the process of detecting the contrast information of the block area of the target image data. However, the contrast detection circuit 6
The image data before being binarized is taken into 3A.

Also in this modification, since the contrast detection circuit 51A is used for both AF control and detection of the characteristic region of the image superimposing processing, the circuit scale does not increase.

FIG. 44 is a diagram showing the characteristic points and corresponding points of the respective pasting parts of the pasting screen in the conventional split photographing type camera.

In this conventional example, the first feature point B is located at an arbitrary position in the pasting area R11 on the screen.
a0 (x1, y1) and the second feature point Bb0 (x2, y2) are x
It is set on the y coordinate. Then, the first corresponding point Bc0 (X3, Y3) and the second corresponding point Bd0 (X4, Y4) corresponding to the area to be superimposed are detected on the XY coordinates. The first feature point Ba0 (x1, y1) and the second feature point B
The inclination angle of b0 (x2, y2) with respect to the y-axis is θ1, and the first corresponding point Bc0 (X3, Y3) and the second corresponding point Bd0 (X4,
The inclination angle of Y4) with respect to the y-axis is θ2. Note that the reference screen and the screen to be pasted together have a predetermined static coordinate deviation in structure other than the positional deviation due to the error of the mirror scan or camera shake, and the xy coordinate system and the XY coordinate system are, for example,

[0096]

## EQU7 ## There is a coordinate shift of (X, Y) = (x, y) +10. Therefore, the values of X and Y are (Equation 7)
Need to be corrected.

In order to obtain the deviation of the bonding in this conventional example, the difference Δx, Δy between the first characteristic point Ba0 (x1, y1) and the first corresponding point Bc0 (X3, Y3) and the second Characteristic point Bb0 (x2, y2) and second corresponding point Bd0 (X4, Y4)
It is necessary to obtain the rotation angle difference Δθ for matching with. That is, after the correction of the above (Formula 7),

[0098]

[Expression 8] Δx = X1 −X3

[0099]

[Formula 9] Δy = Y1−Y3

[0100]

Δθ = arc tan {(x2-x1) / (y2
-Y1)} + arc tan {(X3-X4) / (Y4-Y
3)} In this case, the calculation of the value of Δθ according to (Equation 10) is arc t
Since there are two terms of an, the calculation becomes complicated.

Therefore, in the case of this example, the rotation angle difference Δθ
Apply the shift amount calculation method that can easily calculate. That is, like the two bonding portions M15 and M16 shown in FIG. 12, the first feature point Ba0 (x,
After setting y1), the second feature point Bb0 (x, y2) is set on a line parallel to the y-axis passing through the feature point Ba0. Therefore, the tilt angle θ1 in the above conventional example becomes zero. The first corresponding point Bc0 (X3, on the pasting portion M16 of the screen to be pasted)
Y3) and the second corresponding point Bd0 (X4, Y4) are obtained in the same manner as in the conventional case. The deviation of the first corresponding point and the rotation angle of the second corresponding point are corrected by the above (Formula 7),

[0102]

[Expression 11] Δx = X−X3

[0103]

[Equation 12] Δy = Y1−Y3

[0104]

[Mathematical formula-see original document] [theta] = arc tan {(X4-X3) / (Y4-Y3)} In this case, the value of [Delta] [theta] of the rotation deviation needs only to find one arc tan term, which simplifies the calculation. The accuracy also increases. The circuit scale also becomes smaller.

The above-mentioned calculations of (Equation 11), (Equation 12) and (Equation 13) may be performed by software, but may be performed by a hardware configuration. The arithmetic circuit of FIG. 13 is a block configuration diagram of the shift arithmetic circuit 21a for performing the arithmetic operation, and is composed of four subtractors and an arc tan (a /
It is composed of a ROM table in which the value of b) is written. The shift amount calculation circuit 21a is incorporated in the CPU 21 shown in FIG. Then, the output Δx, Δ
y and θ are output to the first memory controller 16 and the image data in the frame buffer 4 are pasted together.

Next, a modified example of the search processing of the corresponding area of the screen to be pasted corresponding to the contrast state of the characteristic area of the reference screen will be described.

The block sum detection processing in the conventional example shown in FIG. 45 is performed in the pasting screen in which the value of the cumulative addition value Σ of the image data 0 and 1 in the characteristic area of the reference screen, for example, the area of 4 columns × 4 rows, is the same. The corresponding block area above is designated as the corresponding area. In the case of this figure, the cumulative addition value Σ of the characteristic region B21 on the pasting portion M21 on the reference screen side is 8. Then, when the cumulative addition value Σ of the block areas B22, B23, B24 of the pasted screen side pasting section M22 is checked and it is 0, 16, 8, the block area whose cumulative addition value Σ is 8 B24 will be designated as the corresponding area.

In the example of FIG. 46, the pattern matching method of the area is applied, and the block area where the pattern of the image data matches is designated as the corresponding area. In this case, the block area having a pattern that matches the image pattern of the characteristic area B21 specifies the block area B27 of B25, B26, and B27 as the corresponding area.

However, in the case of the block summation method shown in FIG. 45, even if the cumulative addition values Σ match, different image patterns may be designated. Figure 4
As shown in the image data of the characteristic region of No. 7 and the corresponding region, as the region where the cumulative addition value Σ coincides with the characteristic region B31 where the cumulative addition value Σ is 4, the region B where the pattern matches
Block region B3 corresponding to 32 but having a different pattern
3, B34 and B35 are designated as the corresponding areas.

Further, in the pattern matching method shown in FIG. 46, as shown in the screens of the characteristic region and the corresponding region of FIG. 48, the screen M32 to be pasted on the reference screen M31 is rotated and photographed. Even though the characteristic region B36 originally corresponds to the block region, it is determined that the characteristic region B36 does not correspond to the characteristic region B36 because the pattern is relatively inclined.

Therefore, what is proposed as a modified example of the search processing of the corresponding area for the characteristic area in the camera of the above-described embodiment is the processing that uses both the block summation method and the pattern matching method.

In this processing, the bonding screen M52 is bonded to the reference screen M51 shown in FIG.
It is assumed that the combined one shooting screen M50 shown in (B) of FIG. The circuit configuration for executing this processing is substantially the same as the block configuration diagram around the detection circuit shown in FIG. 4, but the detection circuit itself has a block sum detection circuit 18A and a pattern matching detection circuit as shown in FIGS. The difference is that it is composed of 18B.

In this processing, first, the image data of the pasting section R51 of the reference screen M51 and the overlapping pasting section R52 of the pasting screen M52 are fetched into the boundary buffer 17. Then, the block sum detection circuit 18A built in the detection circuit 18 of FIG. 15 obtains the cumulative addition value Σ of the characteristic region B51, and the block region having the cumulative addition value Σ corresponding to the value is searched to find the corresponding region. B
52 is designated, and the shift amount including the rotation component θ is detected.

Subsequently, as shown in the block diagram around the detection circuit shown in FIG. 16, a pasting screen of the frame buffer 4 via the first memory controller 16 based on the deviation amount including the detected rotation component. M52 is rewritten to generate a correction screen M53. After that, the image data of the pasting R53 of the correction screen M53 is further taken into the boundary buffer 17. Then, the bonding portion R of the reference screen
The pattern matching detection between 51 and the bonding section R53 is performed by the pattern matching detection circuit 18B built in the detection circuit 18.

In this case, since the correction screen M53 is subjected to the rotation correction, the block area in which the patterns completely correspond on the cumulative addition value Σ is designated as the corresponding area.

As described above, according to this modification, it is possible to search for a block area in which the cumulative addition value Σ matches and also in which the pattern matches, and to search for a corresponding area that completely corresponds to the characteristic area. Then, as shown in FIG. 47, it is possible to exclude from the corresponding area a block area in which the patterns do not match even if the cumulative addition values Σ match.

Next, another modified example of the search processing of the corresponding area of the screen to be pasted corresponding to the contrast state of the characteristic area of the reference screen will be described.

FIG. 49 is a diagram showing the detection state of the corresponding area detection processing in the conventional example, and shows the first and second characteristic areas A and B when the reference screen M61 and the bonding screen M62 are bonded together. The detected first and second corresponding regions C,
It is the figure which showed D. In this conventional detection process,
A corresponding area C that is moved by Δx and Δy with respect to the characteristic area A is detected, and further, a corresponding area D corresponding to the characteristic area B is detected.
Is detected, and the rotation component θ of the bonded screen is detected.

FIG. 50 is a diagram showing a state in which the reference screen M61 and the pasting screen M62 are pasted together. The pasting screen M62 is moved by the movement amounts Δx and Δy, and further, the relative angle is changed. Rotate by θ and stick together. In this pasted state, the second characteristic region B and the second corresponding region D do not match because the feature region distance AB and the corresponding region distance CD do not match. In this processing, it is possible that the bonding is performed in the state of such a corresponding area detection error.

Therefore, this modification is proposed as the corresponding area detection processing in which the above-mentioned problems do not occur. As shown in FIG. 17, the block diagram around the detection circuit in this process is substantially the same as the block diagram of FIG. 4, but the detection circuit 18 is different.

That is, in the detection processing in this detection circuit,
As shown in the flowchart of FIG. 21, first, the allowable value Δd of the difference of the cumulative addition value Σ of the allowable limit recognized as the corresponding area in the first search is set to 5. That is, if the difference in the cumulative addition value Σ between the characteristic region and the block region is within 5, the block region is treated as the corresponding region. This value Δd is gradually decreased when an appropriate corresponding area cannot be obtained. For example, as shown in the search state diagram of FIG. 18, when the cumulative addition value Σ of the characteristic region B55 is 100, the allowable difference Δd at the first stage
5, the block areas B56, B57, B58 and the like having the cumulative addition value Σ of 80, 95, 99 are compared, and the area B57 is recognized as a search allowable object as a corresponding area. In the processing step of the next step described later, the tolerance Δ
When d is set to 4, only the area B58 is recognized as a search permission target as a corresponding area.

Then, in the flowchart of FIG. 21, p indicating the block area position for searching the first corresponding area is reset to 0. Hereinafter, as shown in FIG. 20, in the bonding section R56, the target area is sequentially changed from Co to Cp while incrementing p, and the cumulative addition value of the block area Cp to the characteristic area A is changed. It is checked whether the difference of Σ is within the value Δd, in this case, within 5. If it is within 5, the block area Cp is set as the first corresponding area.

Further, q indicating the block area position for searching the second corresponding area is reset to 0. Below, FIG.
As shown in 0, in the bonding section R56, the block area D is sequentially incremented from Do by Dq while incrementing q.
The target area is changed to check whether the difference between the cumulative addition value Σ of the characteristic area B and the block area Dq is within a value Δd, that is, within 5 in this case. If it is within 5, the block area Dq is set as the second corresponding area.

Then, as shown in FIG. 19, the distance between the first and second characteristic regions A and B is compared with the distance between the first and second corresponding regions C and D detected by the above processing. To do. When the distances between the two match, the first and second characteristic regions A and B
Assuming that the first and second corresponding areas C and D corresponding to are detected, the shift amount calculation is executed.

When the distances do not match, the tolerance difference value Δd of the cumulative addition value Σ is decremented, and the difference between the cumulative addition value Σ and the block area Cp is within the above value Δd. And whether the difference between the cumulative addition value Σ and the block area Dq is within the value Δd is restarted.

When a matching corresponding area cannot be obtained in this way, the value of the allowable difference is gradually decremented, and the search is repeated in the area where the difference of the cumulative addition value Σ is small.

Even if the allowable difference value Δd of the cumulative addition value Σ becomes 0, and if the corresponding corresponding areas C and D cannot be detected, error processing such as sounding an alarm indicating that the corresponding area cannot be detected is performed. To do.

In addition to the corresponding area detecting process of the modified example shown in the flowchart of FIG. 21, another corresponding area detecting process will be described.

This processing is executed when the first and second characteristic region intervals do not match the detected first and second corresponding region intervals, as shown in the pasting section of the divided photographing screen of FIG. Assuming that the selection itself of the characteristic area of the screen as the reference is not proper, the AF and AE are redone and the shutter speed,
Further, the threshold value or the like at the time of the binarization process is changed, and the selection of the characteristic region and the detection of the corresponding region are redone in that state.

FIG. 22 is a flowchart of the characteristic region / corresponding region detection processing of this modification. First, the reference screen and the bonding screen obtained by performing the AF and AE processing and the binarization processing are written in the frame buffer, and the bonding portion is captured in the boundary buffer. After selecting the characteristic regions A and B, the corresponding regions C and D are detected. Therefore, the distances between the characteristic regions A and B and the distances between the corresponding regions C and D are compared, and if they match, the detected characteristic regions A and B and the corresponding regions C and D have high compatibility. Then, the shift amount is calculated. If they do not match, the constants such as AF and AE processing, the threshold level in the binarization processing, or
The shutter speed and the like are reset, the image is taken, and the image data is written again. Then, the characteristic region is set and the corresponding region is detected.

In addition to the corresponding area detection processing of the modified example shown in the flow chart of FIG. 21, another corresponding area detection processing will be described.

This processing is shown in the flow chart of FIG. 23. The first and second characteristic region intervals and the detected first and second correspondences as shown in the pasting section of the divided photographing screen of FIG. If the area intervals do not match, the zoom magnification is changed, and the corresponding area detection processing is performed based on the changed captured image data.

Further, when no matching corresponding area can be detected by the above-mentioned detection processing, the detection processing of the flowcharts shown in FIGS.

Based on the apparatus of the above-described embodiment or its modification, an apparatus having the following structure is proposed as an image handling apparatus. That is, one image handling device arranges unit areas having predetermined dimensions on the screen and defining the shape and size as units for video signal level detection along a predetermined direction. The unit included in the block area at each stop position when successively moving while forming an overlapping part before and after moving the block area along a predetermined direction An image handling apparatus having a functional unit for evaluating the degree of image contrast in a block area at each stop position by obtaining the sum of video signal levels for each area, and a predetermined number of the block areas. Cumulative addition means for cumulatively adding the video signal level values for each unit area, and the video signal level for each unit area of the block area by the cumulative addition means. And holding means provided to hold the accumulated value of the adding means for obtaining the sum of the value held in said holding means,
Retaining to follow the vertical or horizontal direction sequential movement of the block area and update the value held in a part of the holding means to a value newly obtained by the cumulative addition means in response to the sequential movement. It is characterized by comprising a value updating means.

According to this image handling apparatus, a calculation for evaluating the degree of contrast of an image in the block area by sequentially moving the block area to be evaluated can be executed very quickly, and high-definition image data can be handled quickly. It will be possible.

Another one of the image handling devices includes a first functional unit for identifying whether the image is in the contrast state and evaluating whether the image is in the focus state, and the partially overlapping one area. The above-mentioned image and the image related to the region of the ground are related to the overlapping region, and the region set as at least one characteristic portion in the overlapping region related to at least the one region corresponds to the above. An image handling apparatus having a second function unit for performing a bonding process so as to properly overlap with a region recognized as one corresponding part in the overlapping region related to When the setting is performed, a means is provided for selecting the characteristic portion from the area identified by the first functional portion as having a contrast of a predetermined level or more.

According to this image handling device, the AF means and the means for setting the characteristic portion are partially overlapped and therefore simplified.

Still another one of the image handling devices has an image relating to the one area that partially overlaps and an image relating to the other area in at least the one area with respect to the overlapping area. The first corresponding portion and the second corresponding portion in the overlapping area related to the other area to which the areas set as the first characteristic portion and the second characteristic portion in the overlapping area correspond to these, respectively. An image handling apparatus having a functional unit for detecting a motion so as to be properly overlapped with a region recognized as a unit, the first characteristic unit and the second characteristic unit.
When setting the characteristic portions of the above-mentioned, each means is provided with a means for setting one of the coordinate positions in the Cartesian coordinate system to represent the coordinate position so that they are the same.

According to this image handling device, the detection accuracy (recognition accuracy) and the detection speed are improved, and the circuit scale for calculation is reduced.

In still another image handling apparatus, an image related to the one area and an image related to the other area, which partially overlap each other, may be processed according to at least the one area related to the overlapping area. A first feature and a second in the overlapping region
So that the regions set as the characteristic parts of the above-mentioned regions properly overlap with the regions recognized as the first corresponding part and the second corresponding part in the overlapping region related to the above other regions. An image handling apparatus having a functional unit for processing, the first corresponding unit and / or the second corresponding unit.
The first identification operation of recognizing the corresponding portion of the above is whether or not the accumulated values of the video signal levels in the portion to be compared with the first characteristic portion and / or the second characteristic portion are equal. And a second identifying operation for performing an operation of matching the relative positions of the image of the one area and the image of the other area, and then performing a second identifying operation as to whether or not the patterns in the portion are the same. It is characterized by comprising.

According to this image handling device, the detection accuracy (recognition accuracy) is improved.

In still another image handling device, an image related to the one area and a image related to the other area, which partially overlap each other, may be processed according to at least the one area related to the overlapping area. A first feature and a second in the overlapping region
So that the regions set as the characteristic parts of the above-mentioned regions properly overlap with the regions recognized as the first corresponding part and the second corresponding part in the overlapping region related to the above other regions. An image handling apparatus having a functional unit for processing, wherein a distance between regions recognized as the first corresponding unit and the second corresponding unit is set as the first characteristic unit and the second characteristic unit. When it is discriminated that the distance between the regions is different from the predetermined allowable range, the above first
Setting operation by changing the initial condition for the characteristic part and / or the second characteristic part, the first corresponding part and / or the second
The recognition operation by changing the initial condition for the corresponding portion of, the electronic image magnification variable operation for equalizing the two different distances, or two or more operations of the above three operations are sequentially executed. It is characterized by comprising means for performing the operation.

[0143]

As described above, the image handling apparatus of the present invention has the advantage that a high-contrast area can be quickly selected as the representative point area on the target image, and highly accurate motion detection can be performed.

[Brief description of drawings]

FIG. 1 is a block configuration diagram showing a configuration of a camera which is an image handling apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram showing a split screen of a shooting screen in the camera of FIG.

3A and 3B show a pasted state of a split screen of a shooting screen in the camera of FIG. 1, FIG. 3A is a split screen, and FIG.

4 is a detailed block configuration diagram around a detection circuit in the camera of FIG.

5 is a more detailed block diagram of a detection circuit in the camera of FIG.

6 is a time chart of a detection process of a detection circuit in the camera of FIG.

7 is a diagram showing a moving state of a search block area in the detection processing of the detection circuit in the camera of FIG. 1;

FIG. 8 is a diagram showing a comparison of the number of searches for detecting a characteristic region in the camera of FIG. 1 and a camera separately defined.

9 is a diagram showing a pasted state of divided screens in the camera of FIG. 1;

FIG. 10 is a diagram showing how to combine the divided screens in the camera of FIG. 1;

FIG. 11 is a diagram showing how to combine the divided screens in the camera of FIG. 1;

FIG. 12 is a diagram showing a characteristic region and a corresponding region of a split screen in the camera of FIG.

13 is a diagram showing an example of an addition circuit of a detection circuit in the camera of FIG.

FIG. 14 is a diagram showing a screen pasting state of a modified example of the detection processing in the camera of FIG. 1;

15 is a part of a block configuration diagram of a circuit of a modified example of the detection circuit in the camera of FIG.

16 is a part of a block configuration diagram of a circuit of a modified example of the detection circuit in the camera of FIG.

17 is a block configuration diagram of a circuit of another modified example of the detection circuit in the camera of FIG.

18 is a diagram showing a characteristic region search state of a modified example of the detection processing in the camera of FIG. 1;

FIG. 19 is a diagram showing a characteristic region and corresponding regions of screens to be combined in a modified example of the detection processing in the camera of FIG.

20 is a diagram showing a moving state of a block area for detecting a corresponding area in a modified example of the detection processing in the camera of FIG. 1;

21 is a flowchart of a modified example of the detection process in the camera of FIG.

22 is a flowchart of another modified example of the detection processing in the camera of FIG.

FIG. 23 is a flowchart of yet another modified example of the detection processing in the camera of FIG.

FIG. 24 is a block configuration diagram of a motion detection unit of the image handling apparatus of one embodiment of the present invention.

FIG. 25 is a time chart of the contrast data fetching process of the motion detector of FIG. 24.

FIG. 26 is a block configuration diagram of a contrast detection circuit of the motion detection unit of FIG. 24.

FIG. 27 is a diagram showing divided block areas for contrast detection in the motion detecting section of FIG. 24.

28 is a diagram showing a high-contrast search state in a divided block in a modified example of the detection processing of the motion detection unit of FIG. 24.

29 is a diagram showing a high-contrast search state in a divided block in another modification of the detection processing of the motion detection unit of FIG. 24.

FIG. 30 is a block configuration diagram of a motion detection unit of another example of an image handling apparatus related to the apparatus of the present invention.

31 is a block diagram of a contrast detection circuit of the motion detection unit shown in FIG. 30.

32 is a diagram showing image data of a block area applied to the contrast detection processing of FIG. 31.

[Fig. 33] Fig. 33 is a block configuration diagram of a motion detection unit of an image handling apparatus of still another example related to the apparatus of the present invention.

FIG. 34 is a diagram showing a division state of a detection target screen in the detection processing of the motion detection unit of FIG. 33.

35 is a block configuration diagram of a contrast detection circuit of the motion detection unit of FIG. 33.

[Fig. 36] Fig. 36 is a block configuration diagram of a motion detection unit of an image handling apparatus of still another example related to the apparatus of the present invention.

[Fig. 37] Two superposing parts (bonding part) of the bonding screen in the image handling device according to one concept.
It is a figure which shows the screen of.

FIG. 38 is an enlarged view of a block area applied to the characteristic area detection of the combining process in the image handling device according to one concept.

[Fig. 39] Fig. 39 is an enlarged view of pixels in a block area applied to contrast information detection for searching a characteristic area of a combining process in an image handling apparatus according to one concept.

[Fig. 40] Fig. 40 is a block configuration diagram of an image data cumulative addition value calculation circuit for a bonding process of an image handling device according to one concept.

41 is a time chart of the cumulative addition value calculation circuit of the image handling apparatus of FIG. 40.

FIG. 42 is a block configuration diagram of an image data cumulative addition value calculation circuit for a bonding process of an image handling apparatus according to another concept.

43 is a time chart of a cumulative addition value calculation circuit of the image handling apparatus of FIG. 42.

[Fig. 44] Fig. 44 is a diagram illustrating a characteristic region and a corresponding region in a bonding process of an image handling device according to another concept.

FIG. 45 is a diagram showing a state in which a block area corresponding to a characteristic area in the combining process of the image handling apparatus according to another concept is searched.

FIG. 46 is a diagram showing a state in which a block area corresponding to a characteristic area in a bonding process of an image handling device according to another concept is searched.

FIG. 47 is a diagram showing a state of searching a block area corresponding to a characteristic area in a bonding process of an image handling apparatus according to another concept.

FIG. 48 is a diagram showing a state of searching a block area corresponding to a characteristic area in a bonding process of an image handling apparatus according to another concept.

[Fig. 49] Fig. 49 is a diagram illustrating a characteristic region and a corresponding region in the bonding process of the image handling device according to another concept.

50 is a diagram showing a bonding state in a bonding process based on the concept of FIG. 49 described above.

[Fig. 51] Fig. 51 is a diagram illustrating a target screen and a reference screen to be superimposed in the motion detection process of the image handling device according to another concept.

[Explanation of symbols]

18a1 adder (cumulative addition means) 18a5 adder (addition means) 19 second memory controller (register holding / updating means) Ba first characteristic region (first characteristic portion) Bb second characteristic region (second characteristic portion) Bc First corresponding area (first corresponding portion) Bd Second corresponding area (second corresponding portion) B55, Bf, Bg Characteristic area (characteristic portion) P, Pk Block area

Claims (1)

[Claims] 1. An image related to the one area and an image related to the other area that partially overlap each other are at least one of the overlap area related to the overlap area. A function unit for performing a bonding process such that the area set as one feature section is properly overlapped with the area recognized as one corresponding section in the overlapping area of the other areas that should correspond to these areas. An image handling apparatus having, wherein when setting the characteristic part, an area to be set in itself is divided into a plurality of blocks larger than the characteristic part, and the characteristic part is set in the block. A first discriminating operation of whether or not a predetermined first requirement such as an image contrast value corresponding to the setting is satisfied is executed as a relatively coarse area discriminating operation in each block unit, and then the first discriminating operation is performed. Discrimination Only a block which is discriminated by the operation as satisfying a predetermined requirement corresponding to setting the characteristic part in itself, a predetermined first value such as a contrast value of an image corresponding to setting as the characteristic part in the area in the block is determined. Two
An image handling apparatus, comprising: setting means for executing a second discriminating operation for identifying a small area satisfying the requirement of 1. as a relatively dense area discriminating operation for the in-block area.