JPH0753152B2 - Image data processing system memory data scanning system - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention scans a two-dimensional frame memory that stores an image pattern as pixel data at each address (address) in association with each pixel, Regarding a memory data scanning system of an image measurement processing device that appropriately calls pixel data in the scanning direction to perform image measurement processing, particularly, an image suitable for performing measurement processing of an image pattern obtained using an autorefractometer The present invention relates to a memory data scanning system of a measurement processing device.

(Prior Art) Conventionally, an image measurement processing device for obtaining a shape characteristic of an image pattern has been known. For example, in an autorefractometer, in order to determine the shape characteristics of the ring image,
An image measurement processing device is used for obtaining the positions of at least four coordinate points in the circumferential direction of the ring image. In this conventional image measurement processing device, the video signal from the television camera is used as it is, and as shown in FIG. 6, the slice level SL is determined and the time t ₁ , t ₂ , of the image signal on one horizontal scanning line, Measure t ₃ and t _4, and measure their times t ₁ , t ₂ and
Based on the following equations for t ₃ and t ₄ , the time tx corresponding to the diameter of the ellipse when the ring image is an ellipse is calculated,
The coordinate points of the ring image are detected from the time tx, the coordinate points are obtained at least at four positions in the circumferential direction of the ring image, and the shape characteristic of the ring image is obtained by this.

However, in the image measurement processing device that measures using this video signal as it is, it is limited to scanning in the direction of the horizontal scanning line, and the coordinate point in that direction is dared to have a certain angle direction with respect to the horizontal scanning line as the scanning direction. In order to obtain the above, it is necessary to rotate the measurement optical system so that the scanning direction becomes the direction of the horizontal scanning line, which complicates the structure of the measurement optical system and requires a long measurement time. is there.

(Problems to be solved by the invention) Therefore, in recent years, in an image measurement processing device, a ring image as an image pattern is formed on a two-dimensional image sensor such as a CCD and the image pattern is made to correspond to each pixel. Each address of the frame memory is transferred to a two-dimensional frame memory that stores pixel data at each address, and the physical position of the image pattern corresponds one-to-one to the address of each image that constitutes the two-dimensional image sensor. A memory data scanning method, which is directly related to and measures, is drawing attention.

In this memory data scanning method, the measuring optical system does not have to be rotated when obtaining a coordinate point in a direction having a predetermined angle with respect to the horizontal scanning direction as the scanning direction. However, when performing the measurement process based on the pixel data stored at each address of the frame memory by this scanning method and determining the shape of the ring image by obtaining the coordinate points, the scanning direction is determined in advance and Determine the scan start point and the scan end point, divide the space between the scan start point and the scan end point equally in the scan direction, and specify a plurality of scan points, from the scan start point to the scan end point. By scanning along the scanning direction, the pixel data at each address of the frame memory corresponding to each scanning point is called, the pixel data is treated as a numerical value, and the numerical value is weighted using the distance from the scanning start point. , The weighted data should be averaged to obtain the distance from the scanning start point to the center of gravity of the image in the scanning direction.

For this reason, it is possible to program a program to sequentially instruct the address of the frame memory corresponding to each image of the CCD existing on the straight line in the scanning direction by the microprocessor processor, but sequentially instruct the address by this microprocessor. The scanning method of the memory data has a drawback that the measurement processing time is too long.

A memory data scanning method is also considered in which the scanning direction is specified in advance in a read-only memory ROM as an external memory of the microprocessor and the frame memory address data corresponding to each CCD image existing in the scanning direction is stored. However, since the scanning direction is fixed, there is a drawback that an arbitrary scanning direction cannot be designated.

Furthermore, this memory data scanning method has a problem that the number of scanning lines is fixed.

(Object of the Invention) Therefore, an object of the present invention is to perform an image measurement process capable of appropriately changing the scan direction to a direction different from the horizontal scan direction and the vertical scan direction, and performing a measurement process scan quickly. An object is to provide a memory data scanning method of the device.

(Means for Solving the Problems) A feature of the memory data scanning method of the image measurement processing apparatus according to the present invention is that it is a two-dimensional memory that stores an image pattern to be measured in a direction different from the horizontal scanning direction and the vertical scanning direction. The frame memory of the mold and an address along the measurement direction for each scanning step corresponding to the deviation between the horizontal scanning direction or the vertical scanning direction of the frame memory and the measurement direction to be measured from the scanning start point. It is composed of an addressing unit for designating and a scanning unit for scanning the frame memory along the address designated by the addressing unit.

(Operation) According to the present invention, the frame memory stores the image pattern to be measured in a direction different from the horizontal scanning direction and the vertical scanning direction. The address designating unit designates an address along the measuring direction for each scanning step in correspondence with a deviation between the horizontal scanning direction or the vertical scanning direction of the frame memory and the measuring direction to be measured from the scanning start point. The scanning unit scans the frame memory along the address designated by the address designating unit.

(Embodiment) An embodiment in which an embodiment of a memory data scanning method of an image measuring device according to the present invention is applied to an autorefractometer will be described below with reference to the drawings.

FIG. 1 shows the optical system of the autorefractometer. 1 is a projection optical system for forming a ring-shaped image on the fundus Er of the eye E, 2 is a two-dimensional image sensor for the image of the fundus Er. The imaging optical system for forming on the CCD 3 of 4 is a fixation target optical system for fixing the eye E to be examined in a clouded state, and 5 is an observation optical system for observing the anterior segment Ea of the eye E to be examined. is there.

The projection optical system 1 includes a perforated mirror 6, an infrared LED light source 7,
Relay lens 8, conical prism 9, ring aperture stop
The image forming optical system 2 includes an objective lens 21, a perforated mirror 6, and a relay lens 1.
3, the CCD 3, the fixation target optical system 4 includes an objective lens 21, an infrared transmission visible reflection mirror 14, a relay lens 15, and a fixation target 16. The anterior ocular segment observation optical system 5 includes the objective lens 21 and a half mirror. 20, relay lens 17, image pickup tube 18, monitor TV 19
The ring aperture diaphragm 10 and the fundus Er, the CCD 3 and the fundus Er, the infrared LED light source 7, the perforated mirror 6 and the eye pupil Ep are in a conjugate relationship. The arrangement relationship of each constituent element is as described in Japanese Patent Application No. 60-164829, and therefore its detailed description is omitted.

During the measurement, the image of the anterior segment Ea is always displayed on the screen of the monitor TV 19, the examiner appropriately monitors whether the anterior segment Ea is in a predetermined position, and the subject is The fixation target 16 is clouded to fix the eye E to be inspected. In this state, the infrared light refracted by the conical prism 12 passes through the ring-shaped aperture stop 10 and is reflected by the perforated mirror 6. It reaches the fundus Er through the lens 21 and a primary ring image R ₁ is formed on the fundus Er. The infrared light that has formed the primary ring image R ₁ is the fundus Er.
The secondary ring image R ₂ as an image pattern to be measured in a direction different from the horizontal scanning direction and the vertical direction is reflected on the CCD 3 through the hole 6a of the perforated mirror 6 and reaches the CCD 3. It is formed.

FIG. _{2 is} for explaining the secondary ring image R ₂ formed on the CCD 3. In FIG. 2, the secondary ring image R ₂ is formed on the shaded portion of the CCD 3. Here, the sizes of the primary ring image R ₁ and the secondary ring image R ₂ change according to the refractive index of the eye E, and when the eye E has astigmatism, it is elliptical. It has been known from now on.
A signal charge corresponding to the brightness of each pixel of the ring image R ₂ is deposited on each pixel of the CCD 3. The CCD3 is
The signal charge accumulated in each pixel of the CCD3 driven by the drive circuit 22 of the image measurement processing device shown in FIG.
The pixel data f _j is taken out by the A / D converter 23 and transferred to each address of the frame memory FM, and the pixel data f _j of each pixel of the secondary ring image R ₂ is converted to the frame memory FM.
Is temporarily stored in each address of.

The drive circuit 22 is connected to a control calculation unit MPU that performs measurement control processing to obtain the shape characteristics of the image pattern, and the control calculation unit MPU is a storage circuit 24, a program memory 25, a scanning unit SCU, a display interface 26, The display interface 26 is connected to the accumulator ACU, and the display interface 26 is connected to the display 27. The control operation unit MPU is a program memory 2
5, for exchanging information with the memory circuit 24. Next, the details of the measurement control processing of the control calculation unit MPU will be described in the scanning unit SCU,
It will be described together with the configuration of the accumulation unit ACU.

Here, in the measurement process, as shown in FIG. 2, in order to obtain an ellipse R ₃ , an arbitrary point O (X ₀ , Y ₀ ) is set as the origin, and
A direction that passes through the point O (X ₀ , Y ₀ ) and is at an angle θ with respect to the horizontal scanning direction H is defined as a scanning direction M, and in the direction M,
When the measurement start point P ₁ and the measurement end point P ₂ are determined in advance and a plurality of scanning points are set by equally dividing the range, the relationship between the scanning pitch ΔR and the number N of scanning points is Is.

Here, using the measurement start point P ₁ , the component X _{1 in} the horizontal scanning direction and the component Y ₁ in the vertical scanning direction whose origin is an arbitrary point O (X ₀ , Y ₀ ), P ₁ = (X ₁ , Y It is represented by ₁ ). Similarly, the measurement end point P
₂ is P ₂ = (X ₂ , Y ₂ ) using the horizontal scanning direction distance X ₂ and the vertical scanning direction distance Y ₂ with an arbitrary point O (X ₀ , Y ₀ ) as the origin.
It is represented by.

Therefore, assuming that the pixel pitch of the CCD3 in the horizontal scanning direction is l _N and the pitch in the vertical scanning direction is l _V , X ₂ −X ₁ / l _N = ₂ H ₁ is within the range from the measurement start point P ₁ to the measurement end point P _2. It means the number of CCD3 pixels existing in the horizontal scanning direction. Y ₂ −Y ₁ / l _V = ₂ V ₁ is the measurement start point.
CC existing in the vertical scanning direction within the range from P ₁ to the measurement end point P _2.
It means the number of pixels of D3.

Further, if the scanning pitch in the horizontal scanning direction corresponding to the scanning pitch ΔR is ΔX and the scanning pitch in the vertical scanning direction of the scanning pitch ΔR is ΔY, then ΔX = ΔRcosθ ... ΔY = ΔRsinθ.

Therefore, ΔX / l _N = ΔH is the scanning pitch ΔX in the horizontal scanning direction.
Means the number of CCD3 pixels existing inside, and ΔY / l _V = ΔV
Means the number of pixels of CCD3 existing within the scanning pitch ΔY in the vertical scanning direction. Each pixel of CCD3 and frame memory FM
The addresses (H _i , V _i ) of the frame memory FM corresponding to the i-th scanning point P _i in the scanning direction M of the angle θ correspond to the measurement starting point P ₁ . Assuming that the corresponding address in the horizontal scanning direction is H ₁ and the corresponding address in the vertical scanning direction is V ₁ , (H _i , V _i ) = ([H ₁ + i · ΔH], [V ₁ + i · ΔV])
Given by.

The scanning unit SCU uses this address (H ₁ + i · ΔH, V ₁ + i ·
ΔV) functions as an addressing unit that specifies in cooperation with the control calculation unit MPU, and also as a scanning unit that scans the frame memory along the address specified by the addressing unit. As shown in FIG. 5, the scanning unit SCU includes a step counter (STC) 28, a scanning point number designation counter (PS) 29, a scanning pitch designation register (PR) 30 as a scanning pitch designation unit, and an adder circuit (ADDER). ) 31, address register (AR) 32 as address output unit, comparison circuit 33
It consists of The accumulator ACU consists of a multiplier 34, adders 35, 3
6. It consists of accumulation registers 37 and 38. The address register (AR) 32 and the scanning pitch designation register (PR) 30 are composed of an integer part and a decimal part. The scanning pitch designation register (PR) 30 is loaded with pitch data ΔH and ΔV corresponding to the scanning pitches ΔX and ΔY. The address register (AR) 32 stores the address data (H ₁ , V ₁ ) of the measurement start point P ₁
Is read.

The address register (AR) 32 has an integer part ([H ₁ + i · Δ
H], [V ₁ + i · ΔV]) to output address data designating the address of the frame memory FM. The address of the frame memory FM is the address data H _{i in the} horizontal scanning direction and the address data in the vertical scanning direction. It is determined by the combination of V _i (H _i , V _i ).

The step counter 28 has a function of counting the number i of scanning points in the scanning direction M, and the count data i '
Is input to the comparison circuit 33 and the multiplier 34. Every time the scanning in the scanning direction M is completed, the step counter 28 is reset to "0" by the reset signal from the control arithmetic unit MPU. The adder circuit 31 includes a scanning pitch designation register (PR) for each reading number of scanning points in the scanning direction M.
The pitch data (ΔH, ΔV) held in 30 and the address data (H _i , V _i ) are added including the decimal part, and the added data is used as the next address data (H _i , V _i ) in the address register. It has a function to output to (AR) 32. Adder circuit 31
Serves as an address addition unit that adds pitch data to address data corresponding to the previous scanning point for each scanning step.

The comparison circuit 33 determines whether or not the number of times of reading the scanning points in the scanning direction M matches the number N of scanning points, and the scanning direction M is determined.
It has a function of outputting the scan end signal to the control calculation unit MPU. The pixel data f _i at the address (H _i , V _i ) designated by the address data and the count data i ′ are input to the multiplier 34, and the multiplier 34 receives the count data i ′.
And the pixel data fi are multiplied and output as multiplication data _i ′ · f _i to the accumulation register 37 via the adder 35. The adder 35 has a function of accumulating the multiplication data i ′ · f _i in cooperation with the accumulation register 37, and outputting the accumulated data Σ _i · f _i to the accumulation register 37. The pixel data f _i is read into the adder 36, and the adder 36 cooperates with the accumulation register 38 to accumulate the pixel data f _i and output it as accumulated data Σf _i to the accumulation register 38. The scanning point number designation counter 29 has scanning point number data N'corresponding to the scanning point number N.
Is read.

The control calculation unit MPU uses, as initial data, the address data (H ₁ , V ₁ ) of the measurement starting point P ₁ , the scanning point number data N ′, the pitch data ΔR ′ corresponding to the scanning pitch ΔR, and the initial angle θ.
Has a function of outputting the angle data θ ′ corresponding to, the designated number of scanning directions L, the change angle data Δθ ′ corresponding to the scan direction change angle Δθ, and the scan start code.

Next, the operation of the memory data scanning method according to the present invention will be described while explaining the measurement of the refractive powers S, C and A.

When the measurement is started, the drive circuit 22 drives the CCD 3 under the control of the control calculation unit MPU. As a result, the signal charges accumulated in the CCD 3 are transferred to the frame memory FM via the A / D converter 23 and stored as pixel data f _{j in} each address of the frame memory FM (step S ₁ ). At the same time, initial data is set in the control calculation unit MPU. Then, the control calculation unit MPU sets the change count number j in the scanning direction M to j = 1 (step S ₂ ). This means that the number of changes in the scanning direction M is the first.

Here, the required image pattern is an ellipse indicated by a chain double-dashed line.
As is R _3, based on the coordinate point defined as the intersection of the ellipse R ₃ and the scanning direction M, the number of points for determining the coefficients of the elliptic equation must become at least 4 or more Therefore, the designated number L in the scanning direction is 4
That is all. The change angle data Δθ ′ is appropriately determined based on the designated number L in the scanning direction. If the number N of scanning points in the scanning direction M is large, the measurement accuracy will be improved, but if it is too large, the measurement will take time, and if it is too small, the measurement accuracy will decrease. Since there is no meaning, the pixel number of CCD3 in the scanning direction is set as the maximum limit.

Next, the control calculation unit MPU sets an initial value (step
S ₃ ). That is, a reset signal for setting the read count number i at the scanning point in the scanning direction M to “i = 0” is output to the step counter 28 and the accumulated data Σ
The contents of f _i and Σ _i ′ · f _i are “Σ f _i = 0” and “Σ _i ′ · f _i =
A reset signal is output to the accumulation registers 37 and 38 in order to make it "0".

The control calculation MPU calculates the horizontal scanning pitch ΔH and the vertical scanning pitch ΔV based on the initial data (step S ₄ ). An expression is used for this calculation. When the calculation of the control calculation unit MPU is completed, the pitch data Δ
A scanning start instruction code is added to the beginning of H and ΔV and sent to the scanning unit SCU, and the initial address data (H ₁ , V
₁₎ feeding the scanning point number data N 'to the scanning unit SCU (Step S _5). As a result, the scanning step number data N'is read into the pitch step counter 29. Address data (H ₁ , V ₁ ) is read into the address register (AR) 32. Further, pitch data ΔH and pitch data ΔV are read into the pitch designation register (PR) 30.

Address data is stored in the integer part of the address register (AR) 32.
The integer part of H ₁ and V ₁ is read, and the decimal part is read in the decimal part. The integer part of the pitch designation register (PR) 30 is read with the integer part of the pitch data ΔH and ΔV, and the decimal part is read with the decimal part. Here, when the pitch designation register (PR) 30 receives the scan start instruction code, it starts calculation. That is, the adder circuit 31
Pitch data of pitch specification register (PR) 30 ΔH, Δ
V is sent out, the pitch data ΔH and ΔV are added to the address data (H ₁ , V ₁ ), and the added data is read into the address register (AR) 32 again.

At the same time, the count data “+1” is input to the scanning number counting circuit 28. By inputting this count data, the count number i of the step counter 28 becomes “i =
1 ”. Count data i of the read count number i of the scanning point 'is read in a comparator circuit 33 and the multiplier 34 (step S _6). At that time, the address register (AR)
The previous address data (H ₁ , V ₁ ) held in 32 is
Sent out in the frame memory FM (step S _7). As a result, the pixel data f _i stored at the designated address in the frame memory FM is sent to the multiplier 34 (step S ₈ ).

The multiplier 34 multiplies the pixel data f _i and the count data i ′. The multiplication data i ′ · f _i is read into the accumulation register 37 via the adder 35. The adder 37 is an accumulation register 37
Then, the previous accumulated data held in and the newly transmitted accumulated data are added and sent to the accumulation register 37. Also,
The adder 36 adds the pixel data f _i to the previous accumulated data.
As a result, Σ _i ′ · f _i and Σ f _i are accumulated (step S ₉ ). The accumulation result is repeated until the count number i becomes i = N (step S _10). When the count number i is smaller than the scanning point number N is a "+1" is added to the count number in the step S ₁₁ proceeds to repeat the subsequent steps S _6. The comparison circuit 33 performs the comparison of i = N.
The comparison result is read into the control calculation unit MPU, and as a result, the calculation data Σ _i ′ · f _i and Σf _i are fetched into the control calculation unit MPU.

The control operation unit MPU determines the distance Rθ to the coordinate point in the θ direction, the horizontal distance Xθ to the coordinate point, and the vertical direction to the coordinate point based on the operation data Σ _i ′ · f _i and Σf _i . The distance Yθ is calculated and stored (step S ₁₂ ). Then, the control calculation unit MPU determines the change count number j in the scanning direction.
Determines whether j = L (step S ₁₃ ). _i
= When not L, the process proceeds to step S _14, the j "+
1 ”is added. As a result, the scanning direction change angle Δθ is added to the angle θ, and the scanning direction M in the θ + Δθ direction is designated (step S ₁₅ ). Thereafter, the measurement of the scanning direction is started the process proceeds to Step S _3.

when i = L, the process proceeds to step S _16. Step S
_{In 16} , the coefficient A, based on the calculation formula shown below,
B and C are required.

Ax ² + By ² + Cxy = 1 ... That is, in FIG. 2, assuming that the horizontal scanning direction is the X axis, the major axis of the ellipse R ₃ is a, the minor axis is b, and the angle φ of the major axis a with respect to the X axis is , Angle φ corresponds to the astigmatic axis,
Since the major axis a corresponds to the degree of refraction of the strong principal meridian of astigmatism and the size of the ellipse corresponds to the spherical power, the coefficient A based on the following general formula,
If B and C are obtained, the refractive powers S, C and A can be obtained.

That is, if at least four coordinate points are detected and their coordinate values are obtained, A, B, and C are obtained from the equations by the least squares method, and if a, b, and φ are obtained from the equations, the refraction degree is obtained ( step S _17).

Then, the process proceeds to step S _18. In step S _18, it is determined whether the end of measurement or not. If the measurement has not been completed, the process proceeds to step S ₁ . When the measurement is completed, the control calculation unit MPU displays the measurement result on the display 27.

Although the embodiment has been described above, the present invention is not limited to this, and includes the following.

Although the scanning point number designation register (PS) 29 and the comparison circuit 33 are provided in the present invention, the step counter (STC) 28 may be a subtraction counter instead of providing them. In this case, N · Σf _i − as accumulated data
Σi ′ · f _i will be used.

The determination of the end of scanning at the scanning point in the scanning direction can be made based on the address data in the address register becoming larger than a predetermined value.

Adder circuit (ADDER) 31 is used for pitch specification register (PR) 3
The part corresponding to the code part of 0 may be used as the addition / subtraction counter.

(Effect of the invention) Since the memory data scanning method of the image measurement processing apparatus according to the present invention is configured as described above, even if the number of scanning directions is increased, the hardware is not changed and any direction is set. There is an effect that the address of the frame memory existing in the scanning direction can be quickly specified. Further, before performing the arithmetic processing for obtaining the shape characteristic of the image pattern, the pixel data read from each address of the frame memory is multiplied by the accumulating section to obtain the related data for each scanning direction. , The image pattern can be quickly calculated. Further, since the address designation is performed by calculation including the fractional part, even when the address in the sequential scanning direction is designated, the accumulated error of the address designation can be avoided.

[Brief description of drawings]

FIG. 1 is a diagram showing an optical system of the autorefractometer when the memory data scanning system of the image measurement processing apparatus according to the present invention is applied to the autorefractometer, and FIG. 2 is an image formed on a CCD of the autorefractometer. FIG. 3 is a block diagram of a memory data scanning system of the image measurement processing apparatus according to the present invention, and FIG. 4 is a diagram showing a main configuration of the memory data scanning system block circuit. FIG. 5 is a flow chart for explaining a procedure for obtaining a ring image using the image measurement processing device, and FIG. 6 is a diagram for explaining image measurement processing of a conventional image measurement processing device. 3 …… Image sensor 28 …… Step counter (STC) 29 …… Scanning point number designation counter (PS) 30 …… Pitch designation register (PR) 31 …… Addition circuit (ADDER) 32 …… Address register (AR) 34 …… Multiplier 35,36 …… Adding circuit 37,38 …… Accumulation register MPU …… Control operation unit SCU …… Scanning unit ACU …… Accumulation unit M …… Scanning direction R ₂ …… Secondary ring image (Image pattern) ΔR …… scanning pitch H _i , V _i …… address data θ …… initial angle Δθ …… scanning direction change angle

Claims (4)

[Claims] 1. A two-dimensional frame memory for storing an image pattern to be measured in a direction different from a horizontal scanning direction and a vertical scanning direction, and a horizontal scanning direction or a vertical scanning direction of the frame memory from a scanning start point. And an address designating section for designating an address along the measuring direction for each scanning step in correspondence with the deviation between the measuring direction to be measured and the frame memory along the address designated by the address designating section. A memory data scanning method of an image measurement processing device, which is configured by a scanning unit that scans the. 2. An address output section for outputting address data in the horizontal scanning direction and address data in the vertical scanning direction, and a scanning pitch specification for holding scanning pitch data in the horizontal scanning direction and the vertical scanning direction. The memory of the image measurement processing apparatus according to claim 1, further comprising: an address addition unit that adds the pitch data to address data corresponding to the previous scanning point for each scanning step. Data scanning method. 3. The address adding unit and the scanning pitch designating unit have an integer part and a decimal part, and obtain address data including the decimal part, and the address adder calculates the address data from the integer part of the frame memory. The memory data scanning method of the image measurement processing apparatus according to claim 2, wherein the memory data scanning method is configured to specify an address. 4. The image pattern is a ring image for obtaining the refractive power of an eye to be inspected, and the frame memory includes:
The memory data scanning method of the image measurement processing apparatus according to claim 1, wherein the ring image formed on the two-dimensional image sensor is transferred.