KR101034282B1 - The Method for Controlling Focus in Image Captured from Multi-focus Objects - Google Patents

Abstract

According to an embodiment of the present invention, an image of an object is obtained from a plurality of objects in which a multifocal point is formed along a predetermined optical axis at predetermined distance intervals along an optical axis, and for each obtained image, N having a predetermined width along a first axis. Sets of ROIs at regular intervals, sets M ROIs having a predetermined width along a second axis at regular intervals, and then coordinates of the first and second outermost edges on the first axis of the ROI among the objects (x _p ^-, x _q ⁺⁾ and the second the third and fourth coordinates of the outermost edge of the axis (y _r ^-, y _s ⁺⁾ step and the detected first to fourth outermost edge detecting each the coordinates, but (x _p ^{- -,} x _q ^+, y _r, y _s ⁺⁾ for substituting a predetermined equation, and calculating each of the coordinates of the dynamic focus area defines a dynamic focus area, where, h is the area of interest Where a is the proportional constant and the sharpness of the defined dynamic focus area in each image And calculating each of the distances moved along the optical axis in the image having the highest value of the sharpness of the calculated dynamic focus area as the focal length of the object along the optical axis, respectively. Disclosed is a focus adjustment method in.

According to the present invention, a focal length along an optical axis of an object of interest may be quickly and accurately calculated in an image acquired from a multifocal object.

Multiple Focus, Region Of Interest (ROI), Dynamic Focusing Region (DFR). Field of View (FOV)

Description

The method for controlling focus in image captured from multi-focus objects

The present invention relates to a focus adjustment method in an image obtained from a multifocal object, and more particularly, by applying a defined dynamic focusing region (DFR) to an object of interest in an image obtained from the multifocal object. The present invention relates to a focus adjustment method in an image obtained from a multifocal object capable of quickly and accurately calculating a focal length along an optical axis with respect to an optical axis.

1 is a conceptual diagram of a camera module assembly process including a multifocal object, FIG. 2 is a focus image of each object of an image obtained from the multifocal object in the camera module assembly process of FIG. 1, and FIG. It is a sharpness graph according to the moving distance of the optical axis of the image.

In the process of assembling the camera module, a vacuum gripper 13 is used to align the lens mount 12 equipped with the IR filter 11 with the printed circuit board (not shown). It is necessary to calculate the focal length of the absorbed lens mount 12. The IR filter 11, the lens mount 12, or the vacuum gripper 13 may be defined as a multifocal object.

However, in the image of the vision camera 14 for the camera module assembly process, as shown in FIG. 1, an image having multiple focuses is obtained due to the step difference between the lens mount 12 and the vacuum gripper 13. For example, FIG. 2A illustrates a case where the focus of the vision camera 14 is formed on the vacuum gripper 13, and FIG. 2B illustrates a case where the focus of the vision camera 14 is formed on the lens mount 12. to be.

That is, as shown in FIG. 3, when checking the sharpness graph according to the moving distance with respect to the optical axis of the vision camera 14, it can be seen that two peaks are generated. Here, the first peak I is a case where the focus is formed on the lens mount 12 in FIG. 2B, and the second peak II is the background in FIG. 2A, for example, a vacuum gripper ( The focus is formed at 13).

Therefore, there is an urgent need for acquiring only the image in which the lens mount 12 is focused in the image of the vision camera 14 for the camera module assembly process.

Accordingly, the present invention has been made to solve the above-mentioned conventional problems, and by applying a defined dynamic focus region, a quick and accurate calculation of a focal length along an optical axis of an object of interest in an image obtained from a multifocal object is possible. An object of the present invention is to provide a focusing method in an image obtained from a multifocal object capable of obtaining a clear image of an object of interest.

In order to achieve the above object, according to a preferred embodiment of the present invention, a) acquiring an image of the object at predetermined distance intervals along the optical axis from a plurality of objects in which multiple focal points are formed along a predetermined optical axis And b) in each of the acquired images, N regions of interest having a predetermined width along a first axis are set at regular intervals, and M regions of interest having a predetermined width along a second axis are set at regular intervals. Later, the coordinates (x _p ⁻ , x _q ⁺ ) of the first and second outermost edges on the first axis of the object of interest among the objects and the coordinates y of the third and fourth outermost edges on the second axis _r ^-, y _s ⁺⁾ to a step of detecting each, c) the coordinates (x _p of the first to fourth outermost edge of the detected ^{_{-, x q +, y r}} -, y s +) of a predetermined mathematical Substituting by an equation, each coordinate of the dynamic focus area is calculated to define the dynamic focus area. Where h is the width of the region of interest, a is a proportionality constant, d) calculating the sharpness of the defined dynamic focus region in each of the images, and e) the calculated dynamic Determining the distance moved along the optical axis in the image having the highest value of the sharpness of the focus area as the focal length of the object along the optical axis. to provide.

Preferably, the proportional constant may have a proportionality value of substantially -0.2 to 1.0.

Preferably, the proportional constant is, c1) by substituting the proportional constant value of the range into the equation with a predetermined difference and performing steps a) to d), respectively, to calculate the sharpness of each image for each proportional constant value. And c2) calculating a difference between a first peak value and a second peak value from the calculated sharpness of each image according to the proportional constant value substituted in the equation, using the optical axis as the rotation axis. Rotate each time and repeat steps c1) and c2) to rotate the optical axis as the rotation axis according to the proportional constant value substituted in the equation to determine the first peak value and the second peak value of the sharpness, respectively. The method may further include defining a proportional constant of the highest mean among the differences of the differences.

Preferably, in step c1), the proportional constant may be substituted into the equation by a difference of substantially less than 1/5 of the range in the range of -0.2 to 1.0.

Preferably, in step a), an image of the object may be acquired from the object at intervals of substantially one fifth or less of a viewing range along the optical axis.

Preferably, the steps c1) and c2) may be repeated by rotating the optical axis at intervals of a difference of 1/10 of the expected angle range of the object.

Preferably, the object of interest among the plurality of objects in which the multifocal is formed may be a lens mount.

Preferably, in the step c1), the proportional constant may be substituted into the equation in the range of 0.4 to 0.2 substantially.

Preferably, after step c), the method may further include applying an image processing algorithm within the defined dynamic focus area in each image.

According to the present invention, the dynamic focus region may be applied to quickly and accurately calculate a focus along an optical axis of an object of interest in an image obtained from a multifocal object.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. Here, the focus adjusting method according to an embodiment of the present invention has been described as a focus adjusting method for assembling a lens mount and a printed circuit board in a camera module assembly process, but is not particularly limited thereto, and is an object of interest in a multifocal object. Of course, it can be widely applied to calculate the focal length.

According to an embodiment of the present invention, a method of focusing an image obtained from a multifocal object may include a) acquiring an image of the multifocal object, b) detecting an outermost coordinate of the object of interest in the ROI, and c) dynamically Defining a focus area, d) calculating a sharpness of the dynamic focus area, and e) determining a focal length of the multifocal object.

First, in step a), a predetermined distance along the optical axis from a plurality of objects in which multifocals are formed along an optical axis of an imaging apparatus, for example, a vision camera (not shown), which acquires an image of multi-focused objects. Each image of the multifocal object is captured at intervals.

Each focal image of an object in which multifocals are formed, such as the lens mount 12 and the vacuum gripper 13, which adsorbs the lens mount 12 can be obtained (see FIG. 2). Although only the focus image of the lens mount 12 and the vacuum gripper 13 is shown here, in step a) of the focus adjustment method according to the present embodiment, a certain distance interval along the optical axis, for example, substantially a field of view (FOV; Images of the multifocal object may be acquired through the vision camera at intervals selected to be less than or equal to 1/5 of a view. Here, when the image of the multifocal object is acquired through the vision camera at intervals exceeding 1/5 of the field of view, an error range is significantly increased in calculating a focus on the object of interest 40. It is preferable to obtain images of the multifocal object through the vision camera at intervals selected at a value of 1/5 or less, respectively.

4 is a conceptual diagram for defining a dynamic focus region for an object of interest in a focus control method in an image acquired from a multifocal object according to an embodiment of the present invention.

In step b), in each image obtained in step a), N regions of interest (ROI) having a predetermined width h along the first axis, that is, the X axis are set at regular intervals, and M regions of interest having a predetermined width h along two axes, that is, Y axes, are set at regular intervals.

Then, the multi-focus coordinates of a first and a second outermost edge (edge) of the axis of interest in the target object 40 of the object ^- the third and fourth outermost of the (x _p, x _q ⁺⁾ and a second axial detects _{^{- (, y s + y r}} ) , each coordinate of the edge. Here, the first and second coordinates of the outermost edge of the first axis ^- a (x _p, x _q ⁺⁾ are respectively the smallest coordinate value from the detected edge coordinate every N number of points of interest and the largest coordinate value, a second axis of the third and fourth coordinates of the outermost edge _(r y ^-, y ⁺ _s) are each a smallest coordinate value and a largest coordinate value from the detected edge coordinate for each M number of points of interest.

c) In the step, the first to fourth coordinate of the outermost edge (x _p detected in step b) ^- to a, y ⁺ _s) is substituted by the following equation (1), ^{-, +} x _q, y _r

Each coordinate calculating the ^{(x - -, x +,} y, y +) to dynamically focus area; defined as (DFR dynamic focusing region).

On the other hand, a multi-focus of interest of the subject object (targeted object), for example lens mount 40 has first to fourth coordinate of the outermost edge formed by the (x _p ^-, y _s ^{+ -,} x _q ^+, y _r) Although present in the region to be formed, the formed region does not cover the entire object of interest. Therefore, as expressed in equation (1), a predetermined clearance (margin) (αh) of the first to fourth coordinate of the outermost edge to (x _p ^-, y _s ^{+ -,} x _q ^+, y _r) Will be added. Where h is the width of the region of interest and α is the proportionality constant. In addition, the subscripts p, q, r, s are the number of each edge.

In step d), the sharpness of the dynamic focus area defined in each image in step c) is respectively calculated. That is, as compared with the sharpness calculation time for the entire image of the multifocal object, the sharpness processing time is relatively reduced by calculating the sharpness of only the dynamic focus region in which the object of interest exists inside. Meanwhile, the image of the region except the dynamic focus region of the entire image of the multifocal object may be deleted.

On the other hand, after the above step c), after applying an image processing algorithm (image processing algorithm) in the defined dynamic focus region in each image, it is also possible to calculate the sharpness (sharpness) of the dynamic focus region, respectively.

In step e), the distance moved along the optical axis in the image having the highest value among the sharpness of the dynamic focus area calculated in step d) is determined as the focal length of the multifocal object along the optical axis. That is, the dynamic focus area including the object of interest among the multifocal objects is defined, and the moving distance along the optical axis of the image having the highest sharpness value among the defined dynamic focus areas is determined as the focus along the optical axis of the object of interest. .

FIG. 5 is an image obtained for each rotation angle along an optical axis from a multifocal object. FIG. 6 is a sharpness graph according to a movement distance of an optical axis of an image obtained at a rotation angle of −25 degrees in FIG. 5, and FIG. Sharpness graph according to the moving distance of the optical axis of the image obtained at the rotation angle of 15 degrees, Figure 8 is a sharpness graph according to the moving distance of the optical axis of the image obtained at the rotation angle of -5 degrees in Figure 5, Figure 9 is In Figure 5 is a sharpness graph according to the moving distance of the optical axis of the image obtained at a rotation angle of 5 degrees, Figure 10 is a sharpness graph according to the movement distance of the optical axis of the image obtained at a rotation angle of 15 degrees in Figure 5, Figure 11 is Figure 5 It is a sharpness graph according to the moving distance of the optical axis of the image obtained at a rotation angle of 25 degrees.

5 and 11, the process of determining the proportionality constant assigned to Equation 1 will be described below.

The proportional constant α applied to Equation 1 of this embodiment may have a value of −0.2 to 1.0 substantially.

First, substituting a proportionality value substantially in the range of -0.2 to 1.0 to a value equal to or less than one fifth of the above-described range, for example, substantially equal to a difference of 0.2, respectively, from steps a) to d) The sharpness of each image is calculated for each proportional constant value (see FIGS. 6 to 11) (step c1). Here, if the proportional constant value is substituted into Equation 1 with a difference of more than 1/5 of the above range, the error range for the finally determined proportional constant increases, making it difficult to determine the optimal proportional constant.

That is, by changing the margin value, that is, the proportional constant value, to define the dynamic focus area, the sharpness of the image including the changed object of interest is determined by the proportional constant value (α = -0.2, α = 0.0, α = 0.2, α, respectively). = 0.4, α = 0.6, α = 0.8, α = 1.0).

After the step c1), a difference between the first peak and the second peak is calculated from the sharpness of the image of each dynamic focus region calculated according to the proportional constant value substituted in Equation 1 ( See Table 1) (step c2).

α -0.2 0.0 0.2 0.4 0.6 0.8 1.0 case 1 5.33 4.48 4.33 4.67 3.24 2.04 0.65 case 2 1.40 5.89 3.80 5.30 3.24 3.74 2.08 case 3 3.92 8.66 6.01 7.07 7.45 7.66 7.45 case 4 4.70 2.49 5.54 6.33 6.39 6.16 5.86 case 5 1.81 3.65 3.61 2.65 1.53 0.19 -1.09 case 6 3.18 4.71 6.63 7.60 7.60 4.61 1.99 average 3.39 4.98 4.99 5.60 4.91 4.13 2.83

Finally, steps c1) and c2) are repeatedly performed at intervals of a predetermined angle, for example, a difference of 1/10 of the expected angle range or the expected angle error range of the multifocal object, using the optical axis as the rotation axis. . Here, the expected angle error range is defined as a rotation alignment error range in which the optical axis of the multifocal object aligned substantially perpendicular to the optical axis when the image of the multifocal object is acquired.

In this embodiment, steps c1) and c2) are repeatedly performed by rotating every 5 degrees within the range of -25 degrees to 25 degrees (case1 (-25 degrees), case2 (-15 degrees), case3 (-5 degrees). ), case4 (5 degrees), case5 (15 degrees), case6 (25 degrees)}.

Here, the proportional constant of the highest average among the averages of the difference between the first peak value and the second peak value calculated by rotating the optical axis as the rotation axis according to the proportional constant value substituted in Equation 1 is expressed in Equation 1 below. It is defined as the proportional constant (α) to be substituted. In Table 1, it can be seen that when the proportional constant is 0.4, the average of the difference between the first peak value and the second peak value is 5.60, the highest value.

Therefore, by assigning the proportional constant (α = 0.4) defined as described above to Equation 1, the dynamic focus region of steps a) to e) can be set to quickly and accurately calculate the focal length of the object of interest. On the other hand, in consideration of the calculation error range of the defined proportional constant (α = 0.4), it is preferable that a value in the range of 0.4 to 0.2 is substituted in Equation (1).

1 is a conceptual diagram of a camera module assembly process including a multifocal object.

FIG. 2 is a focus image of each object of an image obtained from a multifocal object in the camera module assembly process of FIG. 1.

3 is a sharpness graph according to a moving distance of an optical axis of the image of FIG. 2.

4 is a conceptual diagram for defining a dynamic focus region for an object of interest in a focus control method in an image acquired from a multifocal object according to an embodiment of the present invention.

5 is an image obtained for each rotation angle along an optical axis from a multifocal object.

FIG. 6 is a sharpness graph according to a moving distance of an optical axis of an image acquired at a rotation angle of −25 degrees in FIG. 5.

FIG. 7 is a sharpness graph according to a moving distance of an optical axis of an image acquired at a rotation angle of −15 degrees in FIG. 5.

FIG. 8 is a sharpness graph according to a moving distance of an optical axis of an image acquired at a rotation angle of −5 degrees in FIG. 5.

9 is a sharpness graph according to a moving distance of an optical axis of an image acquired at a rotation angle of 5 degrees in FIG. 5.

FIG. 10 is a sharpness graph according to a moving distance of an optical axis of an image acquired at a rotation angle of 15 degrees in FIG. 5.

FIG. 11 is a sharpness graph according to a moving distance of an optical axis of an image acquired at a rotation angle of 25 degrees in FIG. 5.

Claims (9)

a) acquiring an image of the object at predetermined distance intervals along the optical axis from a plurality of objects in which multifocals are formed along a predetermined optical axis; b) in each of the acquired images, N areas of interest having a predetermined width are set along the first axis at regular intervals, and M areas of interest having a predetermined width along the second axis are set at regular intervals, first and second coordinates of the outermost edge of the first axis of the interested object of the target object _{^{(x p -, x q +}} ) and the coordinate of the third and fourth outermost edge of the second axis (y _r ^- , y _s ⁺ ), respectively, and a, y ⁺ _s) is substituted by the following equation, ^- c) the coordinates (x _p of the first to fourth outermost edge of the detected ^-, _q ⁺ x, y _r

Each coordinate to calculate the (x ^{- -,} x ^+, y, y ^+), but defined as the dynamic focus area, Where h is the width of the region of interest, a is a proportionality constant, and p, q, r, and s are the number of edges, d) calculating sharpness of the defined dynamic focus area in each of the images, respectively; e) determining the distance traveled along the optical axis in the image having the highest value of the sharpness of the calculated dynamic focus area as the focal length of the multifocal object along the optical axis. To adjust the focus in the image. The method of claim 1, The proportional constant has a proportional constant value of -0.2 to 1.0, the focus control method in the image obtained from the multifocal object. The method of claim 2, The proportional constant is c1) calculating the sharpness of each image by the proportional constant value by substituting the proportional constant value of the range into the equation with a predetermined difference and performing steps a) to d), respectively; c2) calculating a difference between a first peak value and a second peak value in the sharpness of each of the calculated images according to a proportionality value substituted in the equation, The first axis of sharpness calculated by rotating the optical axis as the rotation axis according to the proportional constant value substituted in the equation by repeating steps c1) and c2) by rotating the optical axis at a predetermined angle. And defining a proportional constant of the highest average among the average of the difference between the peak value and the second peak value. The method of claim 3, wherein In the step c1), the proportional constant is substituted in the equation by a difference of 1/5 of the range in the range of -0.2 to 1.0, the focus control method in the image obtained from the multifocal object. delete delete The method of claim 1, The method of adjusting the focus in the image obtained from the multifocal object, characterized in that the object of interest among the plurality of objects that the multi-focus is formed is a lens mount. The method of claim 4, wherein In the step c1), the proportional constant is assigned to the equation in the range of 0.4 to 0.2 focusing method in the image obtained from the multifocal object. The method of claim 1, After step c), further comprising applying an image processing algorithm within the defined dynamic focus area in each image.

Images