JP2006308884A - Automatic focusing device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an automatic focusing device capable of performing high speed and highly precise automatic focusing processing. <P>SOLUTION: A scanning system selection part 34 selects a scanning system of an image. A contrast evaluation value calculation part 26 separates image information into an odd-numbered field and an even-numbered field, calculates contrast evaluation values for every field and combines the two calculated contrast evaluation values when the scanning system is interlace. A contrast evaluation value maximum value judgment part 32 judges maximum in time series change of the contrast evaluation values. A contrast evaluation value maximum Z position calculation part 33 calculates a Z position (focal position) at which contrast of the image information becomes maximum based on a Z position corresponding to the contrast evaluation value judged as maximum. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an autofocus device that captures an image of an observation target, generates image information (image data), and automatically detects a focus position of an observation optical system based on the image information.

  Many microscopes that place an observation object to be inspected or the like on a stage for observation are equipped with an autofocus (AF) function that automatically adjusts the focus. AF processing for realizing the AF function is largely divided into an active method and a passive method. In the active method, a semiconductor laser is irradiated from a microscope onto an observation object, and a focused position is found using reflected light. On the other hand, the passive method finds the in-focus position by using the object detected by the CCD camera or the like.

  Although the active method can perform high-speed processing, it is necessary to incorporate an irradiation light source such as a semiconductor laser and a sensor that detects reflected light, which makes the microscope expensive. On the other hand, the passive method can be realized if the movement control in the Z direction can be performed or the observation image of the observation object can be acquired, so that the microscope is inexpensive, but the stage or the observation system is step-driven in the Z direction. Since the operation of acquiring the image information (image data) after confirming the stop is performed, there is a disadvantage that the processing speed is slow. Therefore, a method for improving the processing speed has been proposed.

  In the technique described in Patent Document 1, a stage or an observation system on which an object to be inspected is moved is moved in the Z direction, the Z position is read from the moving state, and at the same time, image information is acquired, and the degree of contrast of the image is determined. The process of calculating the contrast evaluation value shown is repeated while the stage or the observation system moves within a predetermined range. Thereafter, the AF processing is speeded up by interpolating the plurality of calculated contrast evaluation values to obtain the maximum contrast Z position.

  Further, in the technique described in Patent Document 2, image information is acquired while driving a stage or an observation system on which an object to be inspected is mounted in the Z direction, a contrast evaluation value is calculated, and a contrast peak (maximum) is obtained. I try to find the position and focus. In this method, after a predetermined measurement start point is detected before the actual AF process is started, the AF process is started after recognizing that the stage or the observation system has reached the measurement start point.

  There is also an example in which the AF processing is speeded up by a method that considers the scanning method, which is different from the methods shown in the above two patent documents. Patent Document 3 discloses a method of performing AF processing using field image information under an NTSC format (interlace) condition as an image for calculating a contrast evaluation value, and the observation system moves in the Z direction while moving in the Z direction. While reading the position, the contrast evaluation value of the image information is calculated. Since an image pickup apparatus with a frame rate outputs image information at 30 frames / second, AF processing is also performed at a maximum speed of 30 times / second. In addition, since the field-rate imaging apparatus can output image information at 60 fields / second, AF processing can be realized at a maximum speed of 60 times / second.

In contrast, as described in Patent Document 4, the inventor of the present invention acquires image information while relatively moving a stage or an observation system on which an object to be inspected is placed in the Z direction, and evaluates contrast. A high-speed AF processing method is proposed in which values are calculated and the corresponding image information (contrast maximum image information) is output when the maximum value of the contrast evaluation value is found.
JP 2003-315684 A JP 2003-161875 A Japanese Patent Laid-Open No. 5-288981 JP 2004-62167 A

  In the techniques described in Patent Document 1 and Patent Document 2, the state of acquired image information is not disclosed. For example, in high-magnification observation, vibration generated when moving a stage or an observation system is image information. It is possible to affect Furthermore, when image information is acquired with an interlace scanning camera used in a microscope, the influence of vibration appears in the acquired image information, and as a result, the AF process may fail.

  For example, if vibration occurs in the direction parallel to the scanning line of the image information for a stage or observation system driven in the Z direction, and if an image is acquired with a progressive scan camera, the vibration amplitude is large. However, when an image is acquired with an interlaced scanning camera, a “shift” due to vibration occurs between the odd field and the even field. This “displacement” can be understood even if the amplitude is small, if attention is paid. If there is this “deviation”, the contrast evaluation value is misinterpreted as a high contrast in the calculation of the contrast evaluation value, resulting in a “false focus” that determines that the position is different from the original focus position. become.

  Further, in the technique described in Patent Document 3, AF processing can be realized at a maximum speed of 60 times / second, but changing from a frame to a field is to halve the amount of information inherent in the image. While the processing speed is increased by reducing the amount of information used to calculate the contrast evaluation value, there remains a problem in the accuracy of the in-focus position.

  Further, in the technique described in Patent Document 4, since the change in the Z position does not stop at the in-focus position, when the operator observes, it is necessary to perform the AF processing by the conventional step drive. Further, since there is no mention of the type of scanning method of the camera that acquires image information, as described above, the influence of vibration is reflected in the image in the interlace scanning method, which may cause false focusing.

  Further, conventionally, when the image information is in color, as an observation different from the normal observation, there is a special observation in a state where a filter of a specific color (color filter) is inserted in the imaging system. For example, in the defect repair inspection of industrial substrates (flat panel displays (FPD), semiconductor wafers, etc.), when using a laser repair function that removes defects using an ultraviolet light source with a short wavelength, the optical design is ultraviolet. In order to enhance the contrast at the visible wavelength, the contrast of the imaging system lens (objective lens, etc.) is not clear even if it is in focus. In some cases, a green filter may be attached to improve the contrast of visible image information.

  In this case, since the entire image is green, it is expected that the luminance related to the green component in the image information is increased. At that time, there is a possibility that the luminance is saturated at a bright portion in the image. When the AF processing is performed using the image information of the green component, the luminance saturation affects the calculation of the contrast evaluation value. As a result, it is conceivable that the AF process leads to an erroneous result.

  The present invention has been made in view of the above-described problems, and an object thereof is to provide an autofocus device that can perform high-speed and high-precision autofocus processing.

  The present invention has been made in order to solve the above-described problem, and is an image capturing unit that captures an image of an observation object and generates image information, and a movement that changes a distance between the observation object and the observation optical system. A contrast evaluation value indicating the degree of contrast of the image information in accordance with the control means, the detection means for detecting the distance, the scanning method specifying means for specifying the scanning method of the image, and the specified scanning method of the image; An evaluation value calculating unit that calculates and combines the calculated contrast evaluation values to obtain the contrast evaluation value of the image information, and associates the distance with the contrast evaluation value calculated by the evaluation value calculating unit. A storage unit that stores the determination result, a determination unit that determines a local maximum in the time series change of the contrast evaluation value stored in the storage unit, and the determination unit Based on the distance corresponding to the contrast evaluation value is determined to a large autofocusing apparatus characterized by comprising a distance calculating means for calculating the distance the contrast of the image is maximized.

  Further, in the autofocus device of the present invention, the evaluation value calculation means separates the image information into odd fields and even fields when the scanning method designated by the scanning method designation means is interlaced, The contrast evaluation value is calculated for each field, and the calculated contrast evaluation value is combined to obtain the contrast evaluation value of the image information.

  Further, the present invention provides an imaging unit that images an observation object and generates image information, a movement control unit that changes a distance between the observation object and the observation optical system, and a detection unit that detects the distance. Color information specifying means for specifying color information of the image information used for calculating the contrast evaluation value, and image data composed of the color information specified by the color information specifying means. Evaluation value calculating means for calculating a contrast evaluation value indicating the degree of contrast, storage means for storing the distance and the contrast evaluation value calculated by the evaluation value calculating means in association with each other, and storing in the storage means Determination means for determining a local maximum in the time series change of the contrast evaluation value, and the contrast evaluation value determined to be maximum by the determination means. Based on the distance to an automatic focusing apparatus characterized by comprising a distance calculating means for calculating a distance that the contrast of the image is maximized.

  In the autofocus device according to the aspect of the invention, the movement control unit may be configured such that the movement between the observation object and the observation optical system is based on a time series change of the contrast evaluation value stored in the storage unit. It is characterized by controlling the speed of change of distance.

  In the autofocus device according to the aspect of the invention, the movement control unit may determine whether the evaluation value calculation unit between the observation object and the observation optical system is based on a change in time required for the calculation of the contrast evaluation value. The speed of the change of the distance is controlled.

  In the autofocus device according to the aspect of the invention, the movement control unit may perform the observation based on the depth of focus estimated from the observation optical system and the time required for the evaluation value calculation unit to calculate the contrast evaluation value. An initial speed of the change in the distance between the object and the observation optical system is set.

  In the autofocus device according to the aspect of the invention, the determination unit may determine that the latest contrast evaluation value is equal to or less than a predetermined ratio with respect to the maximum value among the contrast evaluation values stored in the storage unit. Or / and when the contrast evaluation value decreases continuously for a predetermined number of times after the maximum value is calculated, the contrast evaluation value taking the maximum value is determined to be maximal. To do.

  Further, in the autofocus device of the present invention, the distance calculating means calculates the contrast evaluation values acquired before and after the contrast evaluation value determined to be maximal, and the distance corresponding to each of the contrast evaluation values. The distance at which the contrast of the image is maximized is calculated on the basis of an approximation function estimated by interpolating the contrast evaluation value read out from the storage means.

  In the autofocus device of the present invention, timing acquisition means for acquiring a first timing at which the detection means detects the distance and a second timing at which the evaluation value calculation means calculates the contrast evaluation value. In addition, the storage means stores the first timing acquired by the timing acquisition means in association with the distance, and stores the second timing in association with the contrast evaluation value. Features.

  In the autofocus device of the present invention, the color information designating unit designates color information corresponding to a complementary color of the most frequent color from the hue of the image information.

  According to the present invention, since the autofocus process is performed by evaluating the contrast of the acquired image while changing the distance between the observation object and the observation optical system, it is possible to perform a high-speed autofocus process. Is obtained. In addition, when the image scanning method is an interlace method, high-precision autofocus processing is performed by suppressing the influence of image “deviation” due to vibration by calculating and synthesizing contrast evaluation values for each field. The effect that it can be obtained. Furthermore, by calculating the contrast evaluation value using specific color information in the image, there is an effect that it is possible to perform highly accurate autofocus processing with minimal influence of color.

  The best mode for carrying out the present invention will be described below with reference to the drawings. The AF process in the present embodiment is an AF process using image information (image data). For example, an observation target such as an inspection target or an observation system is driven in the Z direction to acquire image information, and an edge or the like is acquired. An evaluation value representing the contrast (contrast evaluation value) is calculated from the image information, and the in-focus position is obtained by moving the inspection object or the observation system to the Z position where the contrast evaluation value is maximized. .

  FIG. 1 is a schematic configuration diagram showing the configuration of the inspection apparatus according to the first embodiment of the present invention. The inspection apparatus according to the present embodiment is equipped with a microscope, and an autofocus (AF) apparatus that performs the autofocus (AF) process of the present invention is mounted along with the microscope. In FIG. 1, the AF device 12 includes a movement control unit 3 that controls movement of the XYZ stage 1 in the Z direction, and a calculation unit 13 that calculates a contrast evaluation value and a focus position. In the present embodiment, the autofocus device according to the present invention will be described using an example in which the autofocus device is applied to a flat panel display (FPD) substrate such as a liquid crystal, a substrate inspection device such as a semiconductor wafer, or a laser repair device. The application of the apparatus is not limited to this.

  On the XYZ stage 1, for example, an FPD glass substrate 2 (observation object) is placed as an object to be inspected. The XYZ stage 1 includes a direct current or alternating current X direction motor, a Y direction motor, and a Z direction motor that move the stage in the X direction, the Y direction, and the Z direction, respectively. Each motor in the X direction, the Y direction, and the Z direction enables constant speed driving and microstep driving. The XYZ stage 1 is driven and controlled in response to a movement instruction output from the movement control unit 3 in the AF device 12. The movement control unit 3 outputs an instruction to move up or down in the Z direction to the XYZ stage 1.

  When receiving an instruction to move in the Z direction, the XYZ stage 1 is raised or lowered in the Z direction continuously at a constant speed or at a predetermined pitch by driving the Z direction motor. At the same time, the movement control unit 3 detects the height information (Z position) of the XYZ stage 1 that is continuously raised or lowered. Above the XYZ stage 1 is provided a CCD camera 5 for obtaining color image information via an observation optical system 4. In the present embodiment, the XYZ stage 1 is driven and moved in the Z direction, but the observation optical system 4 may move in the Z direction. In any case, it is only necessary that the inspection apparatus is configured so that the distance (interval) between the observation object and the observation optical system 4 can be changed.

  The observation optical system 4 has an objective lens 6 and relay lenses 7 and 8 arranged on the optical axis p. The illumination light output from the illumination device 9 enters the half mirror 11 via the illumination relay lens 10. The half mirror 11 is provided on the optical axis p of the observation optical system 4. As a result, the illumination light output from the illumination device 9 is reflected downward by the half mirror 11, and is irradiated onto the FPD glass substrate 2 through the relay lens 7 and the objective lens 6. The CCD camera 5 images the FPD glass substrate 2 placed on the XYZ stage 1 to generate image information and outputs it to the AF device 12.

  Image information output from the CCD camera 5 is composed of luminance information relating to a plurality of colors (hereinafter referred to as color-specific image information). For example, three pieces of information R (red), G (green), and B (blue) are included. It is assumed that it consists of color-specific image information. The output image information is input to the calculation unit 13 in the AF device 12. The calculation unit 13 performs various calculations including an AF process described later. Image information from the CCD camera 5 is input to the display 14 via the calculation unit 13, and the display 14 displays an image based on the image information.

  FIG. 2 is a block diagram showing an internal configuration of the AF apparatus 12 according to the present embodiment. In FIG. 2, the respective configurations of the movement control unit 3 and the calculation unit 13 are illustrated. In the movement control unit 3, the drive motor pulse output unit 21 outputs a pulse for driving the Z-direction motor of the XYZ stage 1 to the XYZ stage 1. In the present embodiment, the number of pulses per unit time is defined as pulses for driving the Z direction motor. By changing the pulse, the speed of the Z direction motor, that is, the moving speed of the XYZ stage 1 in the Z direction can be changed.

  A signal for controlling the pulse is input to the drive motor pulse output unit 21 from the CPU 24, and the drive motor pulse output unit 21 controls a pulse output to the Z direction motor of the XYZ stage 1 based on this signal. The drive motor pulse reading unit 22 reads the current number of pulses of the Z direction motor of the XYZ stage 1. The Z position calculation unit 23 uses the pulse number instruction information from the CPU 24 output to the drive motor pulse output unit 21 or the pulse number read by the drive motor pulse reading unit 22 to observe the FPD glass substrate 2. A current Z position corresponding to the distance to the optical system 4 is calculated.

  In the calculation unit 13, the CPU 24 controls processing of the entire calculation unit 13. The camera image capturing unit 25 captures image information output from the CCD camera 5. The contrast evaluation value calculation unit 26 calculates a contrast evaluation value from the image captured by the camera image capturing unit 25 by a method described later. The elapsed time measurement counter unit 27 is a counter related to elapsed time, and can be initialized at an arbitrary timing.

  The counter reading unit 28 reads the value of the counter of the elapsed time measuring counter unit 27. The value of the counter is calculated based on the Z position calculation timing by the Z position calculation section 23 and the contrast evaluation value calculation timing by the contrast evaluation value calculation section 26 (more specifically, image information used for calculating the contrast evaluation value). This is used as information indicating the timing of acquisition. The Z position storage unit 29 is a memory for storing the Z position calculated by the Z position calculation unit 23. The counter storage unit 30 is a memory for storing the counter value read by the counter reading unit 28. The contrast evaluation value storage unit 31 is a memory for storing the contrast evaluation value calculated by the contrast evaluation value calculation unit 26.

  The contrast evaluation value maximum value determination unit 32 determines whether or not the current maximum value is appropriate as the maximum value using a plurality of contrast evaluation values. The contrast evaluation value maximum Z position calculation unit 33 calculates the Z position (focus position) corresponding to the maximum value when the contrast evaluation value maximum value determination unit 32 determines that the maximum value is appropriate. The scanning method selection unit 34 selects a scanning method for image information output from the CCD camera 5. In the present embodiment, it is assumed that either an interlace scanning method or a progressive scanning method can be selected. The scanning method is detected by, for example, detecting a scanning method from image information or detecting an interval between trigger signals. The color information selection unit 35 selects the color of the image information used by the contrast evaluation value calculation unit 26 based on the image information captured by the camera image capture unit 25. In the present embodiment, one of R (red), G (green), B (blue), C (cyan), M (magenta), Y (yellow), K (grayscale), and A (all colors) is set. It shall be selectable.

  Next, the flow of processing by the above-described configuration is shown. First, the elapsed time measuring counter unit 27 is initialized from the CPU 24 via the counter reading unit 28. At the same time, the CPU 24 instructs the drive motor pulse output unit 21 to start driving the Z direction motor of the XYZ stage 1. That is, the elapsed time measuring counter unit 27 counts the elapsed time from the start of driving the Z-direction motor.

  Subsequently, the CPU 24 instructs the drive motor pulse reading unit 22 to read the current pulse number of the Z direction motor via the Z position calculation unit 23. At the same time, the counter reading unit 28 reads a counter value representing the current elapsed time from the elapsed time measuring counter unit 27. The number of pulses read by the drive motor pulse reading unit 22 is output to the Z position calculation unit 23. The Z position calculation unit 23 calculates a Z position corresponding to the read number of pulses. The Z position calculated by the Z position calculation unit 23 and the counter value read by the counter reading unit 28 are stored in the Z position storage unit 29 and the counter storage unit 30, respectively. At this time, for example, two types of values for the Z position and the contrast evaluation value are prepared as counter values stored in the counter storage unit 30 so that the time relationship between the Z position and the counter value can be matched. The counter value when each calculation of the position and contrast evaluation value is performed is stored so as to be identifiable.

  Subsequently, the CPU 24 instructs the camera image capturing unit 25 to acquire image information via the contrast evaluation value calculating unit 26. At the same time, the counter reading unit 28 reads a counter value indicating the current elapsed time from the elapsed time measuring counter unit 27. The image captured from the CCD camera 5 to the camera image capturing unit 25 is output to the contrast evaluation value calculating unit 26. The contrast evaluation value calculation unit 26 calculates a contrast evaluation value indicating the degree of contrast of the image based on the image information.

  The calculation of the contrast evaluation value corresponds to the color selected by the color information selection unit 35 from the image information in consideration of the scanning method information output from the scanning method selection unit 34 to the contrast evaluation value calculation unit 26. This is performed using color-specific image information (color information). The contrast evaluation value calculated by the contrast evaluation value calculation unit 26 and the counter value read by the counter reading unit 28 are stored in the contrast evaluation value storage unit 31 and the counter storage unit 30, respectively. At this time, as in the case of storing the Z position described above, the time relationship between the contrast evaluation value and the counter value is matched.

  When the contrast evaluation value calculated by the contrast evaluation value calculation unit 26 is stored in the contrast evaluation value storage unit 31, the contrast evaluation value maximum value determination unit 32 includes the contrast evaluation values stored in the contrast evaluation value storage unit 31. Thus, it is determined whether or not there is a reasonable value as the maximum value in the time series change. As a criterion for determining an appropriate maximum value, for example, there is a maximum value among the contrast evaluation values stored in the contrast evaluation value storage unit 31, and when the contrast evaluation value is viewed in time series, the latest contrast evaluation value is obtained. Is less than a predetermined ratio with respect to the maximum value (for example, less than 70%), or when the contrast evaluation value calculated after the time when the maximum value is calculated decreases monotonously three times continuously, etc. It is determined that the contrast evaluation value that takes the maximum value is a reasonable maximum value. Alternatively, when both of the above two conditions are satisfied, it may be determined that the contrast evaluation value that takes the maximum value is a reasonable maximum value. If it is determined that the maximum value of the contrast evaluation value is not appropriate, the determination is performed after obtaining the Z position and calculating the contrast evaluation value again.

  When the contrast evaluation value maximum value determination unit 32 determines the maximum value of the contrast evaluation value (determines that it is appropriate), the contrast evaluation value maximum value determination unit 32 notifies the CPU 24 of the determination of the maximum value. In response to the notification from the contrast evaluation value maximum value determination unit 32, the CPU 24 outputs an instruction to the drive motor pulse output unit 21 to stop the Z-direction motor of the currently driven XYZ stage 1. The instruction to stop the Z direction motor is output from the CPU 24 to the XYZ stage 1 via the drive motor pulse output unit 21. At the same time, the CPU 24 instructs the elapsed time measuring counter unit 27 to stop the operation via the counter reading unit 28. Receiving the instruction, the elapsed time measuring counter unit 27 stops counting.

  Subsequently, the contrast evaluation value maximum value determination unit 32 instructs the contrast evaluation value maximum Z position calculation unit 33 to calculate the Z position (focus position) at which the contrast of the image is maximum. The contrast evaluation value maximum Z position calculation unit 33 receives the instruction, and compares the contrast evaluation value having the maximum value and the plurality of contrast evaluation values acquired before and after the local maximum value with respect to the contrast evaluation value maximum value determination unit 32. Is read out from the contrast evaluation value storage unit 31. In addition, the contrast evaluation value maximum Z position calculation unit 33 reads a counter value corresponding to the timing at which the acquired contrast evaluation value is calculated from the counter storage unit 30, and acquires a plurality of values acquired at timings close to the acquired counter value. The Z position is also read from the Z position storage unit 29.

  The contrast evaluation value maximum Z position calculation unit 33 uses the read value to calculate a Z position where the contrast substantially has a maximum, that is, a focus position, by a method described later. The Z position indicating the in-focus position calculated by the contrast evaluation value maximum Z position calculation unit 33 is output to the CPU 24. The CPU 24 instructs the drive motor pulse output unit 21 to move the XYZ stage 1 to the designated Z position, and drives the Z direction motor of the XYZ stage 1.

  Next, the configuration of the contrast evaluation value calculation unit 26 will be described. FIG. 3 is a block diagram showing an internal configuration of the contrast evaluation value calculation unit 26. FIG. 3 mainly shows the flow of processing until the contrast evaluation value is calculated using the image information output from the camera image capturing unit 25 and the calculation result is output to the contrast evaluation value storage unit 31, for example. The flow of processing related to the acquisition of image information of the camera image capturing unit 25 instructed by the CPU 24 described with reference to FIG. 2 is not particularly shown.

  The image information output from the camera image capturing unit 25 is input to the color information specifying unit 41. The color information specifying unit 41 determines specific color-specific image information used for calculating the contrast evaluation value according to the color selected by the color information selecting unit 35. For example, when the image information is made up of “R”, “G”, and “B” color-specific image information, and “G (green)” is designated by the color information selection unit 35, the color information related to “G” When “M (magenta)” is designated as the image information, the color-specific image information regarding “R” and “G” is determined as the image information used for the calculation of the contrast evaluation value, and the corresponding color-specific image. Information is output.

  In addition, when “K (grayscale)” is designated, the color information designating unit 41 converts the grayscale converted image information into all image information when “A (all colors)” is designated. The image information for each color is determined to be used for calculation of the contrast evaluation value, and the image information as it is input is output to the scan switching unit 42. At the same time, the color information specifying unit 41 outputs, to the contrast evaluation value calculating unit 44, information indicating whether or not calculation such as conversion is necessary for the image information in advance in calculating the actual contrast evaluation value. To do. The information output to the contrast evaluation value calculation unit 44 is, for example, when image information of all colors is used, and when the color information selection unit 35 designates “K (grayscale)”, prior conversion is performed. Similarly, when the color information selection unit 35 designates “A (all colors)”, it is the content that prior conversion is not necessary.

  The image information output from the color information specifying unit 41 is input to the scan switching unit 42. The scan switching unit 42 changes the output destination of the image information according to the scan method instructed from the scan method selecting unit 34. The scan switching unit 42 outputs image information to the field separation unit 43 when the instructed scanning method is interlace, and to the contrast evaluation value calculation unit 44 when the scanning method is progressive.

  The field separation unit 43 operates when the scanning method selection unit 34 specifies that the image information is an interlaced scanning method, and converts the image information output from the scanning switching unit 42 into an odd (Odd) field and an even number (Even). ) Separated into two fields. The image information input to the field separation unit 43 is only the image information of the interlaced scanning method from the contents of the scanning switching unit 42 described above. The field separation unit 43 is provided to perform subsequent calculation of the contrast evaluation value for each field.

  The field-by-field image information separated by the field separation unit 43 and the non-separated progressive scanning image information output from the scan switching unit 42 are input to the contrast evaluation value calculation unit 44. The contrast evaluation value calculation unit 44 performs a calculation for obtaining the contrast evaluation value according to a predetermined method based on the input image information. In the calculation, the contrast evaluation value calculation unit 44 selects and determines either field unit image information or progressive scan method image information as a calculation target in accordance with the scanning method instructed by the scanning method selection unit 34. .

  Further, the contrast evaluation value calculation unit 44 performs processing such as converting image information in advance as necessary based on the information regarding the necessity of image information conversion in advance output from the color information specifying unit 41. . Among these, regarding the conversion of image information, when image information including all colors is input as a target for calculating the contrast evaluation value, a contrast evaluation value is obtained for each color-specific image information, or converted to gray scale. In order to clarify whether the contrast evaluation value is obtained using the image information, the contrast evaluation value calculation unit 44 receives information related to the conversion of the previous image information from the color information specifying unit 41.

  Even if the scanning method of the input image information is different, the same algorithm is used for calculating the contrast evaluation value except that the position referred to by the image information is different. The contrast evaluation value calculated by the contrast evaluation value calculation unit 44 is output to the field-specific evaluation value synthesis unit 45 when it is obtained from the image information in field units, and further subjected to the following processing. When the contrast evaluation value is obtained from the progressive scanning image information, the contrast evaluation value is output to the CPU 24 and the contrast evaluation value storage unit 31 as it is.

The field-by-field evaluation value combining unit 45 operates when interlace is selected as the image information scanning method by the scanning method selection unit 34, and combines the field-by-field contrast evaluation values output from the contrast evaluation value calculation unit 44. The contrast evaluation value of the image information input to the contrast evaluation value calculation unit 26 is calculated. The method for synthesizing the contrast evaluation value of each field is to obtain the contrast evaluation value CV_I of the image information by the expression (1), where CV_I_odd is the contrast evaluation value of the odd (Odd) field and CV_I_even is the contrast evaluation value of the even (Even) field. It is a method.
CV_I = α × CV_odd + (1−α) × CV_I_even (1)

  In the equation (1), α is a weighting coefficient, and 0 ≦ α ≦ 1. By changing this α, the contrast evaluation value CV_I of the image information is only the result calculated from the image information of one of the fields, or the result calculated from the image information of both fields is averaged, etc. Can be controlled. The contrast evaluation value calculated by the equation (1) is output to the CPU 24 and the contrast evaluation value storage unit 31.

  Note that the method of using the calculation results of CV_I_odd and CV_I_even is not limited to the method described above. For example, if a `` deviation '' due to vibration occurs between an odd field and an even field, CV_I_odd and CV_I_even may be significantly different. The calculated contrast evaluation value may not be used for maximal determination. Alternatively, if each of CV_I_odd and CV_I_even is determined to be an abnormal value as a result of comparison with values before and after the time series change of CV_I_odd or CV_I_even, the values may not be used.

Next, a method for calculating the contrast evaluation value by the contrast evaluation value calculation unit 26 will be described. FIG. 4 shows an example of the sampling line pattern used for calculation of the contrast evaluation value in the present embodiment. In FIG. 4, the sampling line pattern is set for the image information Im. Contrast evaluation value calculated by the contrast evaluation value calculating unit 26 is calculated by using the luminance information of the pixels located on the line S ₁ and S ₂ shown in FIG. 4 sampling pattern S ₁ of "×" shape in the center of the screen of (a) the image information Im is set, "×" shape at a plurality of locations, for example four positions shown in FIG. 4 (b) the image information Im sampling pattern S ₂ of is set.

Each sampling pattern S ₁ , S ₂ is used for the following reason. For example, when the FPD glass substrate 2 to be inspected is targeted for defect inspection and the defective portion exists in the center of the screen, when inspecting what the defect itself is, focus on the defect itself. it is necessary to adjust the, by setting the sampling pattern S ₁ in the center of the screen as shown in FIG. 4 (a), AF processing is performed to look at the contrast evaluation value of the defect itself. On the other hand, when inspecting how the defect affects the normal pattern of the FPD glass substrate 2, it is necessary to focus on the normal pattern instead of the defect, as shown in FIG. by setting the sampling pattern S ₂ at a plurality of positions away from the center to, AF processing is performed to look at the contrast evaluation value of the normal pattern. Note sampling pattern is not limited to the shape of the S ₁ and S _2, any screen position, any shape may be set at any number.

Hereinafter, calculation of the contrast evaluation value using the sampling pattern of FIG. 4 will be described. It is assumed that the sampling pattern is composed of N pixel arrays 0 to N−1, and the luminance value of the image information used for calculating the contrast evaluation value is Iv (n) for the nth pixel. Assuming that the luminance value Iv (n) has already been filtered for noise removal, the difference ΔIv (n) with the luminance value Iv (n−1) of the adjacent pixel on the sampling pattern is expressed by equation (2). Ask.
ΔIv (n) = | Iv (n) −Iv (n−1) | ² (2)

  The calculation shown in Expression (2) is performed on each pixel on the sampling pattern, and the value CV_I expressed in Expression (3) is used as the contrast evaluation value.

  The right side of the equation (3) indicates the sum of ΔIv (n) from n = 1 to N−1. Equation (3) relates to one sampling pattern, and when there is a set of two sampling patterns configured in an “x” shape as shown in FIG. The sum of the obtained “CV_I” may be used as the contrast evaluation value of the entire image information. Also, when there is a plurality of “×” type sampling patterns as shown in FIG. 4B, for example, when there are a plurality of “×” type sampling patterns, the weight is changed according to the screen position of the “×” type sampling pattern. Thus, the contrast evaluation value of the entire image information may be obtained.

  Next, AF processing in the present embodiment will be described with reference to FIG. FIG. 5 is a flowchart showing a procedure of AF processing in the present embodiment. In FIG. 5, only a schematic description is given to the processing that has already been described. Based on the instruction from the CPU 24, the elapsed time measuring counter unit 27 initializes the value of the counter (step S101). The drive motor pulse output unit 21 starts driving the Z direction motor provided in the XYZ stage 1 and moves the XYZ stage 1 in the Z direction (step S102). Based on the number of pulses read by the drive motor pulse reading unit 22, the Z position calculation unit 23 calculates the current Z position of the XYZ stage 1, and at the same time, the counter reading unit 28 obtains the current counter value. The acquired Z position and counter value are associated with each other as a set, and are stored in the Z position storage unit 29 and the counter storage unit 30, respectively (step S103).

  The contrast evaluation value calculation unit 26 acquires image information for calculating a contrast evaluation value from the camera image capturing unit 25. At the same time, the CPU 24 obtains the current counter value from the counter reading unit 28 (step S104). The contrast evaluation value calculation unit 26 calculates a contrast evaluation value related to the image information acquired in step S104. The calculated contrast evaluation values are associated with the counter values acquired in step S104 as one set, and are stored in the contrast evaluation value storage unit 31 and the counter storage unit 30, respectively (step S105).

  The contrast evaluation value maximum value determination unit 32 reads the contrast evaluation value calculated so far from the contrast evaluation value storage unit 31, and determines whether or not the maximum value of the current contrast evaluation value is appropriate (step). S106). The criteria for this determination are as described above. The determination result is notified from the contrast evaluation value maximum value determination unit 32 to the CPU 24. The CPU 24 determines which of the following processes should be performed based on the result of step S106 (determination result of whether or not the maximum value of the contrast evaluation value is appropriate) (step S107). If the maximum value is valid, the process proceeds to step S108. If the maximum value is not valid, the process proceeds to step S111.

  When the maximum value of the contrast evaluation value is appropriate, the CPU 24 instructs the drive motor pulse output unit 21 to stop the Z-direction motor currently being driven, and the received drive motor pulse output unit 21 receives the instruction. The Z direction motor is stopped (step S108). The contrast evaluation value maximum Z position calculation unit 33 calculates the Z position, that is, the in-focus position when the contrast of the image is maximum corresponding to the maximum value of the contrast evaluation value, and outputs the calculation result to the CPU 24 ( Step S109). When calculating the in-focus position, the contrast evaluation value maximum Z position calculation unit 33 refers to the contrast evaluation value and the counter value associated with the Z position, and not only the maximum value of the contrast evaluation value but also the Contrast evaluation values calculated before and after are also used. The CPU 24 instructs the drive motor pulse output unit 21 to drive the Z-direction motor. Upon receiving the instruction, the drive motor pulse output unit 21 drives the Z-direction motor, and the XYZ stage 1 is calculated in step S109. Move to the in-focus position (step S110).

  If the maximum value of the contrast evaluation value is not appropriate, the CPU 24 stores the latest contrast evaluation value and the contrast evaluation value corresponding to the time (counter value) one time before the latest. It is read from the unit 31, and it is determined whether or not the change between them is equal to or greater than a predetermined threshold value Th1 (step S111). If the change between the two contrast evaluation values is greater than or equal to the threshold value Th1, the process proceeds to step S112. If the change is less than the threshold value Th1, the process returns to step S103.

  The process in step S111 is a process for enabling the drive speed of the Z-direction motor to be changed in accordance with the temporal change in the contrast evaluation value. The threshold value Th1 varies depending on the definition of the contrast evaluation value and the pattern of the object to be inspected. For example, when the possible range of the contrast evaluation value is the minimum value 0 and the maximum value 10000, the threshold value Th1 is set to +3000. (Contrast evaluation value is increased), and the value is set in consideration of the sufficiently large change in the temporally adjacent state.

When the time change of the contrast evaluation value is equal to or greater than the threshold value Th1, the CPU 24 sets the drive speed of the Z direction motor to the current β times (0 <β <1) (decelerates) the drive motor pulse. The output unit 21 is instructed. Upon receiving the instruction, the drive motor pulse output unit 21 controls the drive speed of the Z-direction motor as described above (step S112). With this process, the Z position corresponding to the image information for calculating the contrast evaluation value can be acquired at a finer interval based on the recognition that the in-focus position is close as the contrast evaluation value increases. It is a process to do. If the process of step S112 is executed many times, the driving speed of the Z-direction motor is slowed down and affects the AF processing time. Therefore, the process is executed a predetermined number of times Q so that the driving speed is not too slow. And For example, when the Z-direction motor drive speed at the start of processing = Vz0 and β = 0.8, the speed Vzq when the process of step S112 is performed q times is expressed by equation (4).
Vzq = Vz0 × β ^q (4)

In order to prevent the left side of the equation (4) from being less than B times Vz0, the number Q of processing is determined as in the equation (5).
Q = Int (LogB / Logβ) (5)
In equation (5), “Int (X)” represents the maximum integer not exceeding X. For example, if ＝= 0.35 when β = 0.8, Q = 4, and the process of step S112 is executed up to four times.

  Next, the method of determining the validity of the maximum value of the contrast evaluation value using the change of the contrast evaluation value, and the method of changing the driving speed of the Z-direction motor of the XYZ stage 1 by the contrast maximum value determination unit 32 of the present embodiment. Will be explained. FIG. 6A is a diagram for explaining a method of determining the validity of the maximum value using the change in the contrast evaluation value, and FIG. 6B is a diagram illustrating the driving speed of the Z-direction motor of the XYZ stage 1. It is a figure for demonstrating the change method. In both figures, the horizontal axis represents the Z position, and the vertical axis represents the contrast evaluation value.

The validity of the maximum value of the contrast evaluation value is determined by the method described above. In FIG. 6A, when the contrast evaluation value “CV_I _max ” when the Z position is “Z _max ” is the maximum value of the contrast evaluation values calculated up to the present time, A threshold value regarding the contrast evaluation value (for example, a value of 70% of the maximum value of the contrast evaluation value in this embodiment) is set.

In the subsequent processing, it is assumed that the contrast evaluation value “CV_I _m ” when the Z position is “Z _m ” is less than the threshold for determining the validity of the maximum value, that is, “CV_I _max × 0.7”. At this time, it is determined that the contrast evaluation value “CV_I _max ” is appropriate as the maximum value, and processing for calculating the in-focus position by stopping the Z-direction motor is performed. As another method for determining the validity, for example, when the maximum value of the contrast evaluation value is calculated and continuously decreased monotonically three times, it is determined that the maximum value is appropriate. In the case of FIG. 6A, when the contrast evaluation value “CV_I _m ” when the Z position is “Z _m ” is calculated, it is determined that the contrast evaluation value “CV_I _max ” is valid as the maximum value. The

The change of the driving speed of the Z direction motor has been described with reference to FIG. In FIG. 6B, it is assumed that the driving speed V _z of the Z direction motor is “X [pps]”. The contrast evaluation value when the Z position is “Z _m ” is “CV_I _m ”, and the calculated contrast evaluation value when the Z position is “Z _{m + 1} ” is “CV_I _{m + 1} ”. Suppose there was. The CPU 24 reads out the contrast evaluation values that are temporally adjacent from the contrast evaluation value storage unit 31, calculates the difference ΔCV_I, and determines whether ΔCV_I is equal to or greater than the threshold value Th1 in step S111 in FIG. Investigate.

Assuming that the difference between the contrast evaluation value “CV_I _m ” and the contrast evaluation value “CV_I _{m + 1} ” is ΔCV_I _m , if ΔCV_I _m is equal to or greater than Th1, the CPU 24 recognizes that it has approached the in-focus position and is in the Z direction. It is determined that the motor driving speed V _{z is set} to ½ of the current speed, that is, “X / 2 [pps]”. Here, “1/2” corresponds to “β” in equation (4). As a result, the interval between the Z positions calculated thereafter is reduced, and the contrast evaluation value, that is, the in-focus position is calculated with higher accuracy than before.

The process shown in FIG. 6 is described on the assumption that the characteristic of the contrast evaluation value is a “mountain” type that exhibits a relatively “steep” characteristic as viewed from the peak value. When the evaluation value characteristic is a gradual “mountain” type, when the difference ΔCV_I _m between the consecutively calculated contrast evaluation values is equal to or greater than the predetermined threshold value, the predetermined number of times continues. In addition, the speed of the Z direction motor may be changed. In addition, even if the contrast evaluation value difference ΔCV_I _m after changing the Z-direction motor drive speed is examined and the contrast evaluation value is calculated a predetermined number of times, it is still determined that the maximum contrast evaluation value is still valid. If not, the contrast evaluation value difference ΔCV_I _m is compared with a predetermined threshold value, and if ΔCV_I _m is less than the threshold value, the driving speed of the Z-direction motor is returned to the state before the change (ie, increased). What should I do?

Next, a method for calculating the in-focus position by the contrast evaluation value maximum Z position calculating unit 33 according to the present embodiment will be described with reference to FIG. In FIG. 7, when it is determined that the contrast evaluation value “CV_I _m ” calculated when the Z position is “Z _m ” is the maximum value, the contrast evaluation value maximum Z position calculation unit 33 displays the maximum value. The Z position and the contrast evaluation value acquired before and after are read from the Z position storage unit 29 and the contrast evaluation value storage unit 31, respectively. The contrast evaluation value maximum Z position calculation unit 33 interpolates the contrast evaluation value using them, and approximates the estimated curve of the contrast evaluation value to a quadratic function by, for example, the least square method, thereby increasing the contrast (contrast evaluation value). The Z position at which is the largest, that is, the focus position is estimated. In FIG. 7, five Z positions (Z _{m−2 to} Z _{m + 2} ) and contrast evaluation values (CV_I _{m−2 to} CV_I _{m + 2} ) are used.

If the relationship between the Z position and the contrast evaluation value is a general quadratic function Y = A × X ² + B × X + C (Z position is expressed as X and contrast evaluation value is expressed as Y), the coefficients A, B, C is estimated. Then, when X is obtained such that the differential function Y ′ = 2 × A × X + B of the quadratic function is 0, the X corresponds to the Z position for focusing. In FIG. 7, the Z position “Z _max ” slightly separated from the Z position “Z _m ” corresponding to the actual contrast evaluation value “CV_I _m ” is estimated as the in-focus position. Thereby, even if the Z position is acquired and the contrast evaluation value is calculated at a rough interval, the in-focus position can be obtained with high accuracy. The approximate function representing the relationship between the Z position and the contrast evaluation value need not be limited to a quadratic function. Further, the method for estimating the coefficient of the approximation function may not be limited to the least square method.

  Next, the association between the Z position and contrast evaluation value and the counter value in this embodiment will be described. FIG. 8 shows the relationship in the storage unit between the counter value in the counter storage unit 30, the contrast evaluation value in the contrast evaluation value storage unit 31, and the Z position in the Z position storage unit 29. Since the Z position and the contrast evaluation value are associated with the counter value and stored in each storage unit, the counter value stored by the counter storage unit 30 corresponds to either the Z position or the contrast evaluation value. It is necessary to identify whether there is. Therefore, as shown in FIG. 8A, the counter storage unit 30 stores the index value corresponding to the value of each counter added to the counter value. The corresponding index value is a value for identifying the counter value associated with the Z position or the contrast evaluation value.

  The Z position stored by the Z position storage unit 29 and the contrast evaluation value stored by the contrast evaluation value storage unit 31 are stored in association with the corresponding index values stored in the counter storage unit 30, respectively. . Thus, the counter values of the corresponding Z position and contrast evaluation value can be managed separately.

  Next, a method for calculating the in-focus position with higher accuracy will be described with reference to FIG. FIG. 9 is a diagram illustrating the transition of the counter value and the change of the Z position and the contrast evaluation value in the present embodiment. The horizontal axis is the counter value, and the vertical axis is the Z position and contrast evaluation value. Ideally, the acquisition of the Z position and the calculation of the contrast evaluation value of the corresponding image (accurately, acquisition of image information for calculation) are performed at exactly the same timing (time), but actually the entire apparatus Due to factors such as the processing load, the Z position acquisition and the image information acquisition are not necessarily performed at the same timing, and a slight time lag may occur between both timings.

  When high-precision AF processing such as AF processing performed in an observation system with a very narrow depth of focus is considered, it is conceivable that the Z position and the time lag in obtaining image information affect the processing accuracy. Therefore, in the present embodiment, as described below, a counter value representing timing is acquired at each timing of acquisition of the Z position and acquisition of image information (calculation of contrast evaluation value), and the influence of the time lag. To reduce as much as possible.

  In FIG. 9, the driving of the Z direction motor is started, the Z position is acquired, the image information is acquired, and the contrast evaluation value is calculated. When acquiring both data, the counter reading unit 28 reads the counter value of the elapsed time measuring counter unit 27. Each acquired or calculated data is stored in a corresponding storage unit with the data structure shown in FIG. This series of processing is continued until it is determined that the maximum value of the contrast evaluation value is appropriate.

If it is determined that the maximum value of the contrast evaluation value is appropriate, the contrast evaluation value maximum Z position calculation unit 33 uses the Z position corresponding to the contrast evaluation value as it is when calculating the in-focus position. First, the “Z position at the time of image information acquisition” is estimated and used using the counter value. For estimation of the Z position at the time of image information acquisition, a counter value corresponding to the Z position immediately after the calculation of the contrast evaluation value is used. The m-th Z position is “Z _m ”, the corresponding counter value is “TC _2m ”, the counter value corresponding to the acquisition of the m-th image information is “TC _{2m + 1} ”, the next (m + 1) th Assuming that the Z position is “Z _{m + 1} ” and the corresponding counter value is “TC _{2m + 2} ”, the Z position “Ev_Z _m ” at the time of image information acquisition is estimated by equation (6).
_{Ev_Z m = {Z m + 1} × (TC 2m + 1 - TC 2m) + Z m × (TC 2m + 2 -TC 2m + 1)} / (TC 2m + 2 - TC 2m) ··· (6)
For a plurality of contrast evaluation values used for calculation of the in-focus position, the Z position is estimated and used by the expression (6), thereby reducing the influence of the time lag in obtaining the Z position and image information, and highly accurate AF processing. It can be performed.

  In this embodiment, image information from a color CCD camera (area sensor) provided in the microscope is handled, but the means for obtaining the image information is not particularly limited and can be applied to a line sensor. In this case, processing equivalent to the calculation method using the sampling pattern shown in FIG. 4 can be realized by using luminance information of all pixels for calculation of the contrast evaluation value. Also, the Z position is obtained based on the Z direction motor pulse of the XYZ stage 1, but instead of the Z position, a pulse value, for example, a value read from the encoder of the motor itself is used as a substitute for the Z position. May be processed.

  As described above, the autofocus device according to the present embodiment acquires the Z position and the image information while moving the stage or the observation system on which the object is arranged in the Z direction, and calculates the contrast evaluation value calculated based on the image information. And the Z position (position corresponding to the distance between the object to be observed and the observation optical system) are stored in association with each other, and whether or not the contrast evaluation value is maximized in the time series change is determined. When it is determined that the Z value is satisfied, the Z position corresponding to the contrast evaluation value that takes the maximum, that is, the in-focus position is calculated. Since the AF processing is performed by evaluating the contrast of the acquired image while reading the Z position during movement in the Z direction, the speed can be increased compared with the AF processing by the conventional driving method in which the driving and stopping are sequentially performed.

  Also, the autofocus device of this embodiment adapts the method of calculating the contrast evaluation value to the image scanning method (progressive / interlace), and divides information (field / frame unit) used for the AF processing. The autofocus apparatus according to the present embodiment calculates the contrast evaluation value for each of the odd field and the even field and synthesizes both, particularly in the case of the interlace method. As a result, in the interlace method, the influence of image “deviation” or “blurring” caused by vibration between the adjacent fields of the image acquired in the environment with vibration is minimized, and the disturbance is prevented. And stable and highly accurate AF processing can be realized.

  In addition, the autofocus apparatus according to the present embodiment calculates the contrast evaluation value using specific color information in the image as necessary in consideration of the color state of the image. As a result, even when a color filter is inserted in the observation system or when the white balance is biased toward a specific color, it is possible to realize a highly accurate AF process with minimal influence of color. it can.

  Further, since the Z position when the contrast evaluation value is small and the time series change is small is away from the in-focus position, the autofocus device of the present embodiment increases the speed of the Z position change, and the contrast evaluation value is When the series change becomes large, it is determined that the in-focus position is close, and the speed of change in the Z position is reduced. Thus, by controlling the speed according to the state of the contrast evaluation value, it is possible to realize high-speed and high-precision AF processing.

  In addition, the autofocus device according to the present embodiment is configured such that the latest contrast evaluation value is equal to or less than a predetermined ratio with respect to the maximum value among the already calculated and stored contrast evaluation values, or / and the maximum value is set. After the calculation, when the contrast evaluation value decreases continuously a predetermined number of times, it is determined that the maximum contrast evaluation value is the maximum. As described above, since the validity of the maximum is determined based on the situation of the decrease in the contrast evaluation value after the contrast evaluation value taking the maximum is calculated, it is possible to realize an AF process with high accuracy.

  In addition, the autofocus device according to the present embodiment is estimated by interpolating the contrast evaluation values using the contrast evaluation values acquired before and after the contrast evaluation value determined to be maximal and the Z position corresponding to each contrast evaluation value. Based on the approximate function, the Z position where the contrast becomes maximum, that is, the in-focus position is calculated. As a result, it is possible to realize AF processing capable of maintaining the same accuracy as that at the time of low-speed driving even in a situation where the contrast evaluation value is calculated at high-speed driving.

  Further, the autofocus device according to the present embodiment associates the timing (or time) at which the Z position is detected with the Z position, and stores the timing (or time) at which the contrast evaluation value is calculated and the contrast evaluation value in association with each other. To do. By associating the respective acquisition timings with the Z position and the contrast evaluation value, even if the acquisition timing of the Z position and the calculation timing of the contrast evaluation value corresponding to the Z position are shifted, the timing difference between the two Since it is possible to determine the maximum of the contrast evaluation value in consideration of the above, it is possible to realize highly accurate AF processing.

  Next, a second embodiment of the present invention will be described. Since the configuration of the autofocus device according to the present embodiment is basically the same as that of the first embodiment, only the points different from the first embodiment will be described. The difference from the first embodiment is the method of changing the driving speed of the Z direction motor of the XYZ stage 1 and the processing in the color information specifying unit 41.

  First, a method for changing the driving speed of the Z-direction motor will be described with reference to FIG. FIG. 10 is a flowchart showing the procedure of AF processing in the present embodiment. The procedure of the AF process is basically the same as the procedure shown in FIG. 5 except for the processes in steps S121 and S122. In the first embodiment, the speed of the Z-direction motor is changed based on the latest change in the contrast evaluation value, but in this embodiment, one contrast evaluation is performed based on the value of the counter indicating the acquisition timing of the image information. The time for calculating the value is checked, and the driving speed of the Z-direction motor is changed according to the time.

  Here, the “initial speed” when starting to drive the Z-direction motor will be described. When the contrast evaluation value is calculated while the Z-direction motor is being driven, how much the XYZ stage 1 moves in the Z direction at the time interval for acquiring contrast evaluation values that are temporally adjacent, that is, the image information acquisition interval (Z How fast the direction motor moves the XYZ stage 1) affects the accuracy of the AF process. Therefore, the depth of focus of the observation system is taken up as an element for setting the speed at the start of driving.

The depth of focus of the observation system refers to the depth of focus determined from the entire lens related to observation, such as the objective lens 6 and the lens of the CCD camera 5. That is, the contrast evaluation value for one piece of image information is calculated while the XYZ stage 1 moves by the distance of the focal depth. The focal depth of the observation system is DOF, the contrast evaluation value calculation time (counter value) for one piece of image information is ΔTC _C , and the initial speed Vz0 at the start of driving the Z-direction motor is given by equation (7).
Vz0 = DOF / ΔTC _C (7)

The calculation time (counter value) ΔTC _C of the contrast evaluation value given by the equation (7) is the time for acquiring one piece of image information when the calculation of the contrast evaluation value is performed in the order of 10 milliseconds or less. If the scanning method of the CCD camera 5 is an interlaced scanning method, the scanning time of one frame image (33 milliseconds) may be given as a guide.

  When the initial speed Vz0 is set, the drive is actually started, and AF processing is performed, it is conceivable that the time interval for acquiring image information increases due to an increase in processing load inside the apparatus. In this case, if the processing is performed while maintaining the speed set by the equation (7), the time interval for calculating the contrast evaluation value is widened, and the accuracy is lowered when the in-focus position is estimated by the approximation function. There is a concern. Therefore, in the present embodiment, in step S121, the CPU 24 obtains a difference between the counter value when the latest image information is acquired and the counter value when the image information one time earlier from the latest is acquired, If this is equal to or greater than the threshold value Th2, the interval for calculating the contrast evaluation value is widened, and it is determined that the accuracy of the AF processing will deteriorate if this is left as it is.

When the difference between the counter values is equal to or greater than the threshold value Th2, the CPU 24 controls to reduce the driving speed of the Z-direction motor in accordance with the difference between the counter values in step S122. For example, with reference to how much the difference between the counter values increases with respect to the contrast evaluation value calculation time (counter value) ΔTC _C given by equation (7), the counter value If the difference is around the scanning time (66 milliseconds) of the image of two frames, the CPU 24 performs control such as reducing the driving speed to the current half.

  Next, processing in the color information specifying unit 41 of the present embodiment will be described with reference to FIG. FIG. 11 shows a method for estimating color information used for calculating a contrast evaluation value based on image information in the present embodiment. In the first embodiment, the color information selection unit 35 selects a color to be used for calculation of the contrast evaluation value, and the color information designation unit 41 extracts image information for each color that is actually required. The color information specifying unit 41 according to the embodiment automatically specifies the image information for each color used for calculating the contrast evaluation value based on the image information. The image information is assumed to be composed of RGB colors.

  The color information specifying unit 41 performs HSL conversion on the image information input from the camera image capturing unit 25 and extracts the hue information (H). Hue is expressed as an angle determined by the position of the standard color wheel, and the range is from 0 ° to 360 °. In the present embodiment, hue information is calculated using B (blue) as a reference (0 ° and 360 °). Subsequently, the color information specifying unit 41 creates a histogram as shown in FIG. 11 for the calculated hue information. In this histogram, the horizontal axis represents hue and the vertical axis represents frequency. Of these, the horizontal axis is divided into six colors (R, G, B, C, M, Y), and further, one color is divided into six sections according to the angle.

  FIG. 11 shows an example of a histogram of hue information obtained from image information of an observation system in which a green color filter is inserted. The color information designating unit 41 counts one frequency for information of one pixel. In the sense of examining the hue tendency of the entire screen, for example, one color information is obtained by obtaining one hue information from the average of four adjacent pixels or nine pixels. One frequency may be counted. In consideration of the input of grayscale image information, it is also possible to determine whether or not to add hue information to the histogram after examining the saturation information (S) required during HSL conversion. Good. When using saturation information, for example, if saturation is represented by 256 gradations (minimum 0 is gray without saturation, maximum 255 is pure color), and saturation is 25 gradations or less, it is calculated in the same place The processed hue information is not added to the histogram.

  Subsequently, the color information specifying unit 41 selects a hue with the highest frequency based on the histogram of the hue information. The hue with the highest frequency is the dominant “color” in the image information. In FIG. 11, the section “Yellow” (yellow) is the section with the highest frequency. This includes factors such as illumination light and lenses in addition to green color filters. In consideration of the input of grayscale image information, a histogram of a section having a frequency less than a certain fixed value may not be used for selection of hue. For example, “effective minimum frequency” parallel to the horizontal axis (hue) in FIG. 11 is set, and those whose frequency is less than the effective minimum frequency are excluded from hue selection targets. Then, the hue with the highest frequency is actually selected from either “Yellow” or “Green”.

  Subsequently, the color information specifying unit 41 determines “complementary color” for “hue” dominant in the image information, and determines this as image information for each color used for calculation of the contrast evaluation value. Luminance information corresponding to “color” dominant in image information tends to be relatively large (bright), and may be saturated near the maximum luminance. Since the luminance information close to saturation cannot reproduce the gradation sufficiently, if the luminance information is used for calculating the contrast evaluation value, the contrast evaluation value cannot be calculated appropriately. Based on this, the opposite “color” = “complementary color” has little luminance information close to saturation with respect to the “color” dominant in the image information. Therefore, if “complementary color” is used, an appropriate contrast evaluation value is obtained. Can be calculated. For example, if the hue having the highest frequency in FIG. 11 is “Yellow”, the complementary color is “Blue”, and therefore image information for each color of B (blue) among the image information of RGB is used. A contrast evaluation value is calculated. For the subsequent calculation of the contrast evaluation value, the same method as in the first embodiment is used.

  In this way, the hue of the image information is examined, and the color-dependent image information consisting of “complementary colors” that are dominant in the image information is used for calculating the contrast evaluation value, thereby minimizing the influence of the screen color. Thus, highly accurate AF processing can be performed. Note that the process of determining “complementary colors” need not be performed every time image information is input. For example, processing is performed only once using the reference inspection target at the stage of setting the inspection apparatus or microscope, or processing is performed only once using image information obtained by imaging the inspection target in the first inspection. The determined color information may be reflected in subsequent inspections. In addition, when the “hue” of the image information differs depending on the type of the object to be inspected and the observation method, information on the inspection conditions may be used. For example, a separate storage unit for color information may be provided so that “complementary color” information corresponding to the inspection condition is stored in the storage unit, and “complementary color” information is assigned from the inspection condition in actual inspection. Good.

  As described above, the autofocus device according to the present embodiment sets the initial speed of the change in the Z position based on the depth of focus estimated from the observation system and the time required to calculate the contrast evaluation value. As a result, it is possible to realize an AF process that can change the Z position at high speed while maintaining the same accuracy as that of low-speed driving.

  In addition, the autofocus device according to the present embodiment controls the speed of change of the Z position based on the time series change of the contrast evaluation value. Even if the calculation time of the contrast evaluation value (the processing time of the contrast evaluation value calculation unit 26) increases due to an increase in the load due to the control of the entire apparatus, Z depends on the degree of change in the calculation time. By changing the speed of the position change, it is possible to realize AF processing capable of maintaining stable accuracy.

  In addition, the autofocus device according to the present embodiment designates color information corresponding to the complementary color of the most frequent dominant color from the hue of the image information, and uses it to calculate the contrast evaluation value. If the dominant color that occupies the most in the entire image is defined as the image dominant color, the color information corresponding to the complementary color of the image dominant color is not biased toward the brighter frequency. This indicates that “saturation (overexposure)” that affects the contrast evaluation value of the image is the smallest. Therefore, by using the color information that has the most information that is effective for calculating the contrast evaluation value, it is possible to calculate the contrast evaluation value of an appropriate image and to realize highly accurate AF processing.

  As described above, the embodiments of the present invention have been described in detail with reference to the drawings. However, the specific configuration is not limited to these embodiments, and includes design changes and the like without departing from the gist of the present invention. .

It is a schematic block diagram which shows the structure of the test | inspection apparatus by the 1st Embodiment of this invention. 1 is a block diagram illustrating a configuration of an AF apparatus according to a first embodiment of the present invention. It is a block diagram which shows the structure of the contrast evaluation value calculation part by the 1st Embodiment of this invention. It is a reference figure showing an example of a sampling line pattern used for calculation of contrast evaluation values in a 1st embodiment of the present invention. It is a flowchart which shows the procedure of operation | movement of AF apparatus by the 1st Embodiment of this invention. FIG. 5 is a reference diagram for explaining a method for determining the validity of the maximum value of the contrast evaluation value and a method for changing the driving speed of the Z-direction motor in the first embodiment of the present invention. It is a reference figure for demonstrating the calculation method of the focus position in the 1st Embodiment of this invention. FIG. 6 is a reference diagram for explaining association between a Z position, a contrast evaluation value, and a counter value in the first embodiment of the present invention. It is a reference figure showing change of a value of a counter in a 1st embodiment of the present invention, and change of a Z position and contrast evaluation value. It is a flowchart which shows the procedure of operation | movement of AF apparatus by the 2nd Embodiment of this invention. It is a reference diagram for demonstrating the process in the color information designation | designated part 41 in the 2nd Embodiment of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... XYZ stage, 2 ... FPD glass substrate, 3 ... Movement control part, 4 ... Observation optical system, 5 ... CCD camera, 6 ... Objective lens, 7, 8, ... Relay lens, 9 ... lighting device, 10 ... lighting relay lens, 11 ... half mirror, 12 ... AF device, 13 ... calculation unit, 14 ... display, 21 ... Drive motor pulse output unit, 22 ... Drive motor pulse reading unit, 23 ... Z position calculation unit, 24 ... CPU, 25 ... Camera image capturing unit, 26 ... Contrast evaluation value Calculation unit, 27... Elapsed time measurement counter unit, 28... Counter reading unit, 29... Z position storage unit, 30. ... Maximum contrast evaluation value , 33 ... Contrast evaluation value maximum Z position calculation part, 34 ... Scanning method selection part, 35 ... Color information selection part, 41 ... Color information designation part, 42 ... Scan switching part, 43... Field separation unit, 44... Contrast evaluation value calculation unit, 45.

Claims (10)

An imaging means for imaging an observation object and generating image information;
Movement control means for changing the distance between the observation object and the observation optical system;
Detecting means for detecting the distance;
Scanning method designating means for designating an image scanning method;
In accordance with the designated image scanning method, a contrast evaluation value indicating the degree of contrast of the image information is calculated, and the calculated contrast evaluation value is combined to be used as the contrast evaluation value of the image information. Means,
Storage means for storing the distance and the contrast evaluation value calculated by the evaluation value calculation means in association with each other;
Determination means for determining a local maximum in the time series change of the contrast evaluation value stored in the storage means;
Distance calculation means for calculating a distance at which the contrast of the image is maximum based on the distance corresponding to the contrast evaluation value determined to be maximum by the determination means;
An autofocus device characterized by comprising:   The evaluation value calculation means calculates the contrast evaluation value for each field by separating the image information into odd fields and even fields when the scanning method designated by the scanning method designation means is interlaced. The autofocus device according to claim 1, wherein the calculated contrast evaluation values are combined to obtain the contrast evaluation value of the image information. An imaging means for imaging an observation object and generating image information;
Movement control means for changing the distance between the observation object and the observation optical system;
Detecting means for detecting the distance;
Color information specifying means for specifying color information of the image information used for calculation of the contrast evaluation value;
Evaluation value calculating means for calculating a contrast evaluation value indicating the degree of contrast of the image information, using image data composed of the color information specified by the color information specifying means;
Storage means for storing the distance and the contrast evaluation value calculated by the evaluation value calculation means in association with each other;
Determination means for determining a local maximum in the time series change of the contrast evaluation value stored in the storage means;
Distance calculation means for calculating a distance at which the contrast of the image is maximum based on the distance corresponding to the contrast evaluation value determined to be maximum by the determination means;
An autofocus device characterized by comprising:   The movement control means controls the speed of change in the distance between the observation object and the observation optical system based on a time series change in the contrast evaluation value stored in the storage means. The autofocus device according to any one of claims 1 to 3, wherein the autofocus device is characterized.   The movement control means controls the speed of change in the distance between the observation object and the observation optical system based on a change in time required for the evaluation value calculation means to calculate the contrast evaluation value. The autofocus device according to any one of claims 1 to 3, wherein:   The movement control means is based on the depth of focus estimated from the observation optical system and the time required for the evaluation value calculation means to calculate the contrast evaluation value, between the observation object and the observation optical system. 6. The autofocus device according to claim 1, wherein an initial speed of a change in the distance is set.   The determination means calculates when the latest contrast evaluation value is equal to or less than a predetermined ratio with respect to the maximum value of the contrast evaluation values stored in the storage means, or / and the maximum value is calculated. 7. The method according to claim 1, wherein when the contrast evaluation value decreases continuously a predetermined number of times, the contrast evaluation value that takes the maximum value is determined to be maximal. The autofocus device according to any of the above items.   The distance calculating unit reads the contrast evaluation values acquired before and after the contrast evaluation value determined to be maximal and the distance corresponding to each of the contrast evaluation values from the storage unit, and reads the read contrast evaluation values 8. The autofocus device according to claim 1, wherein the distance at which the contrast of the image is maximized is calculated based on an approximate function estimated by interpolating values. .   The storage means further comprises timing acquisition means for acquiring a first timing at which the detection means has detected the distance and a second timing at which the evaluation value calculation means has calculated the contrast evaluation value. The first timing acquired by the acquisition unit is stored in association with the distance, and the second timing is stored in association with the contrast evaluation value. The autofocus device according to any one of the items. 10. The auto according to claim 3, wherein the color information designating unit designates color information corresponding to a complementary color of the most frequent color from the hue of the image information. Focus device.

Images