JP2543077B2 - Optical system for focus detection - Google Patents

Description

The present invention relates to a focus detection optical system.

[Problems to be Solved by Prior Art and Invention]

The image formed by the projection lens is divided by a re-imaging optical system consisting of a pupil projection lens and a pair of re-imaging lenses to be re-formed on the sensor, and the light intensity distributions of these two images are compared and calculated. Conventionally, many focus detection optical systems have been proposed in which the distance between two images is obtained to obtain a focus shift amount.

As such an example, "a device for photoelectrically determining the position of a clear surface of an image" described in JP-A-52-82419, and "a focus detection device" described in JP-A-55-118019. , Etc., all of these devices merely show the basic configuration of the focus detection system.

As an example focusing on the problem of the illuminance distribution of the re-imaging optical system, especially the reduction of the peripheral light amount, there is a "camera focus detection device" described in JP-A-60-32014. The problem remains that the amount of light in different parts of the image is different. FIG. 8 schematically shows the illuminance distribution on the sensor surface for a subject having a uniform luminance distribution. As shown in the figure, there is a deviation between the center of the illuminance distribution and the centers C ₁ and C ₂ of the image. This is called the illuminance distribution center shift. This is a factor that reduces the focusing accuracy.

Focusing on the problem of the illuminance distribution center deviation,
There is a "focus detection device" described in Japanese Patent Application Laid-Open No. 55-155331, which corrects the illuminance distribution center shift by adding an auxiliary lens between the pupil projection lens and the re-imaging lens.
An "optical system of a focus detection device" described in Japanese Patent Laid-Open No. 6225715 has a similar structure.

In the optical system described in JP-A-62-25715,
We have achieved correction of the illuminance distribution center deviation, but co
The problem of peripheral light reduction due to the s4 power law is still unsolved. If there is a decrease in the amount of peripheral light, there is an adverse effect that the detection accuracy deteriorates especially during defocusing.

That is, not only is the decrease in the amount of light an error factor, but the contrast of the secondary image is decreased during defocusing, so the influence of the error is further increased. That is, since an accurate focus shift amount cannot be obtained during large blurring, it becomes necessary to repeat the focus detection operation several times until reaching the focus point, resulting in a delay in the focus speed. For example, FIG. 9 shows a state of the secondary image of the dark line chart at the time of the rear focus. However, the reliability of the correlation calculation is deteriorated because the identity of the images is broken. Therefore, a re-imaging optical system in which the amount of peripheral light is reduced as little as possible is desirable, and for example, a re-imaging optical system as shown in FIG. 10 is desirable.

An optical system described in JP-A-60-32014 uses a plano-convex lens as the shape of the re-imaging lens to improve the amount of peripheral light, but this simply widens the distance between the front principal point and the diaphragm. It is a simple measure, and even if it is a biconvex lens, the same effect can be obtained by moving the diaphragm a little away from the lens and moving it toward the front, so it is hard to say that it is a drastic measure. Further, in this optical system, as shown in FIG. 8, there is a deviation between the center of the illuminance distribution and the center of the image. Further, an auxiliary lens for correcting this is added to JP-A-55-155331 and JP-A-62-
As described in Japanese Patent No. 25715, all of them add a new lens, which causes an increase in parts cost and assembly cost. For example, in the optical system disclosed in Japanese Patent Laid-Open No. 60-32014, as shown in FIG. 11, the optical components are the pupil projection lens 1 and the re-imaging lens 2, whereas in the optical system disclosed in Japanese Patent Laid-Open No. 62-32014. In the optical system described in the -25715 publication, one auxiliary lens 3 is further added as shown in FIG.

In view of the above problems, the present invention has good focus detection accuracy, particularly excellent defocus detection accuracy at the time of large blurring, and an optical system for focus detection that does not increase component costs and assembly costs. The purpose is to provide.

[Means and Actions for Solving Problems]

A focus detection optical system according to the present invention is located near a planned image forming surface of an image forming lens, and has a pupil projection lens for transmitting an exit pupil of the image forming lens, and a single lens formed by the image forming lens. In a focus detection optical system including a pair of re-imaging lenses that divides an image and re-forms it on a sensor, the front surfaces of the pair of re-imaging lenses include a correction refracting surface coaxial with the pupil projection lens. The rear surface is made up of a pair of image forming working surfaces in which convex surfaces are formed in parallel on the sensor side, so that the unevenness of the illuminance distribution can be corrected without increasing the number of parts.

That is, it is very important to correct the non-uniformity of the illuminance distribution of the two images on the sensor surface in order to ensure the focusing accuracy. In other words, it is necessary to take appropriate measures against the problems of center deviation due to illuminance and reduction of peripheral light intensity.

The deviation of the illuminance distribution center is corrected by adding one auxiliary lens, but this leads to an increase in the number of parts. Therefore, by integrating the auxiliary lens unit 3'and the re-imaging lens unit 2'as shown in FIG. 1, not only the desired target of correcting the deviation of the illuminance distribution center but also the number of parts is increased. Effective.

The decrease in the amount of peripheral light is mainly due to the cos ⁴ power law, that is, the decrease due to cos ⁴ ω when the angle of view is ω.
Assuming that the refracting power of the re-imaging acting portion is ψs and the effective image height on the sensor is I, cos ⁴ ω = (1-sin ² ω) ² (1-tan ² ω) ² = (1-ψs ² I ² ) ² .

It is desirable that the amount of reduction in the peripheral light amount be at least 10% or less, and considering deterioration due to aberrations, vignetting (vignetting), etc., it is desirable that cos ⁴ ω be at least 0.92 or more. Therefore, 0.92 <cos ⁴ ω (1-ψs ² I ² ) ² ∴ψsI <0.202.

By satisfying this condition, not only the detection accuracy at the time of focusing can be improved, but also an accurate amount of defocus can be detected even at the time of defocusing, and the focusing drive can be speeded up.

Here, it will be described in detail that the correction of the illuminance distribution center shift is particularly effective for improving the focusing accuracy, and the improvement of the peripheral light amount is effective for detecting the focus shift amount during defocus.

If the two sensor rows are the A and B rows and the i-th pixel output is a _i , b _i , the correlation calculation formula is, for example, Then, if s is changed and the correlation calculation is repeated, a curve of F (s) is obtained. FIG. 2 shows a graph of the pixel output of two images and the correlation calculation value with respect to the dark line chart when the re-imaging optical system has no unevenness of the illuminance distribution. If the two images are out of phase by δ, the position of the F (s) curve having the extreme value is also δ, and the extreme value itself is very close to 0, so the images are also displaced by δ. Is clearly understood,
From this, the defocus amount can be calculated.

Therefore, FIG. 3 shows a graph of the pixel output of two images and the correlation calculation value at the time of focusing in the case where the illuminance distribution center shifts. When the illuminance distribution center shifts, the shapes of A and B images differ as shown in the figure. Figure 4 shows the images A and B superimposed, and it can be seen that they do not completely overlap. That is, even if the correlation calculation value becomes an extreme value, F (s) does not become 0 because a component corresponding to the area of the hatched portion in FIG. 4 remains.
In addition, the shaded area in FIG. 4 is completely symmetrical because the subject is located at the perfect center of the distance measuring frame and the quantization error is 0. Can't be in. That is, this symmetry is a necessary condition for the extreme value of the correlation calculation to be at the position of 0, but it is not realistic as described above. That is, as shown in the graph of F (s) in FIG. 3, the error Δδ always remains.

In addition, the phase of the two images is
FIG. 5 shows a state of a graph of the pixel output of two images and the correlation calculation value in the case where the two images are displaced and are in the rear focus. A as shown
When comparing the output ranges of a _{K1 to} a _{K2 for} the image and b _{K1 + S to} b _{K2 + S} for the B image, the A and B images do not completely match. That is, for the same reason as in FIGS. 3 and 4, the extreme value of the correlation calculation value does not become 0, and the error Δδ ′ remains.

As described above, the above-mentioned error occurs due to the nonuniformity of the illuminance distribution, and the focus detection optical system of the present invention focuses on the illuminance distribution of this secondary image to improve its uniformity, thereby improving the focusing ability. Has been achieved.

〔Example〕

Hereinafter, the present invention will be described in detail based on the illustrated embodiments.

FIG. 6 shows an embodiment of the focus detecting optical system according to the present invention. 11 is an image forming lens such as a photographing lens of a single-lens reflex camera, 12 is a planned image forming surface of the image forming lens 11, 13 is an exit pupil of the image forming lens 11 and is projected on a diaphragm 15 immediately before the re-imaging lens 14. A pupil projection lens, 16 is a photoelectric conversion element array.

FIG. 7 shows the re-imaging lens 14 and the photoelectric conversion element array 16 shown in FIG.
Is enlarged. The front surface of the re-imaging lens 14 is a correction refracting surface for making a ray passing through the center of the diaphragm 15 parallel to the optical axis O as shown in the drawing, and is coaxial with the optical axis O. The rear surfaces 14b ₁ and 14b ₂ of the re-imaging lens 14 are configured as a pair of convex surfaces facing the photoelectric conversion element array 16 side, and are re-imaging action surfaces for forming a secondary image on the photoelectric conversion element array 16. is there.

A numerical example of the re-imaging lens 14 shown in FIG. 7 is shown in Table 1 below. However, the refracting power ψs of the re-imaging action surfaces 14b ₁ and 14b ₂ is ψs
Standardized as = 1. In Table 1, R is the radius of curvature, D is the air gap and lens thickness, Nd is the refractive index for the d-line, νd is the Abbe number, and I is the effective image height on the photoelectric conversion element array 16.

The re-imaging lens 14 is integrally molded of plastic,
As shown in FIG. 7, a flat surface portion is provided on the side portion on the front surface side to serve as a receiving surface for the diaphragm 15.

The distance Σ between the centers of the re-imaging lenses 14b ₁ and 14b ₂ is Σ = 2I
= 0.348. The decrease in illuminance due to the cos4 power law is about 6.0% at the maximum image height.

 Other numerical sequences are shown in Table 2 below.

It should be noted that Σ = 2I = 0.274, and the decrease in illuminance due to the cos4 power law is about 3.7% at the maximum image height.

As described above, according to the focus detection optical system of the present invention, the correction refractive surface
The illuminance distribution center shift is corrected by 14a, and the refracting powers of the re-imaging action surfaces 14b ₁ and 14b ₂ are set to ψs, and the photoelectric conversion element array 16
When the effective image height is set to I, φsI <0.2, so that the peripheral light amount reduction is extremely small. Therefore, since the illuminance distribution is highly uniform, the focusing accuracy is significantly improved.
Further, since the correction refracting surface 14a and the re-imaging action surfaces 14b ₁ and 14b ₂ are integrated as the re-imaging lens 14, the number of parts does not increase, and as a result, the parts cost and the assembly cost do not increase. I'm done.

〔The invention's effect〕

As described above, it is important that the focus detection optical system according to the present invention has good focusing accuracy, can detect an accurate focus deviation amount even at the time of defocusing, and does not increase the component cost and the assembly cost. It has many advantages.

[Brief description of drawings]

FIG. 1 is a diagram showing a basic configuration of a focus detection optical system according to the present invention, and FIG. 2 is a two-image image for a dark line chart at the time of non-focus when the re-imaging optical system has no unevenness of illuminance distribution. FIG. 3 is a graph showing a pixel output and a correlation calculation value graph, FIG. 3 is a graph showing a pixel output and a correlation calculation value graph of two images when the illuminance distribution center shift is present, and FIG. 4 is the two images of FIG. FIG. 5 is a diagram showing the pixel output of two images and a correlation calculation value when out-of-focus when the peripheral light amount decrease is large, and FIG. 6 is an embodiment of the focus detection optical system of the present invention. FIG. 7 is an enlarged view of a main part of FIG. 6, FIG. 8 is a view showing an illuminance distribution on a sensor surface for an object having a uniform luminance distribution in the conventional example, and FIG. 9 is a conventional view. Fig. 10 is a diagram showing the illuminance distribution of the secondary image of the dark line chart when the subject is out of focus in Fig. 10. Shows an illuminance distribution of the secondary image of the dark line chart when unfocused cases have, FIG. 11 and FIG. 12 is a diagram showing the configuration of each prior art and other related art. 11 ... Imaging lens, 12 ... Planned imaging surface, 13 ... Pupil projection lens, 14 ... Re-imaging lens, 14a ... Correction refraction surface, 14b ₁ ,
14b ₂ ...... Re-imaging surface, 15 ...... diaphragm, 16 ...... Photoelectric conversion element array.

Claims (2)

(57) [Claims] 1. A pupil projection lens located near a planned image forming surface of an image forming lens and transmitting an exit pupil of the image forming lens, and a primary image formed by the image forming lens is divided and placed on a sensor. In a focus detection optical system consisting of a pair of re-imaging lenses to be re-formed into, and an aperture stop, the front surface of the pair of re-imaging lenses comprises a convex lens surface coaxial with the pupil projection lens, and a rear surface. Is a pair of image forming surfaces in which convex surfaces are formed in parallel on the sensor side, and the aperture stop is provided on the front surface of the convex lens surface. 2. When the refractive power of the pair of image forming surfaces is ψ _S and the effective image height of the secondary image on the sensor is I, then ψ _S I <0.2 is satisfied. ) The optical system for focus detection according to [1].