JP5169641B2 - Image analysis apparatus and image analysis program - Google Patents

Description

  The present invention relates to a technique for adjusting the brightness of an image photographed by a camera, and in particular, the camera image so that the brightness of the photographed image is suitable for recognition processing of characters and patterns. The present invention relates to an image analysis apparatus and an image analysis program that automatically adjusts the amount of light and automatically corrects the overall brightness of an image stored and transmitted so that a person can easily see it.

  2. Description of the Related Art In recent years, image processing systems that install an imaging device (camera) outdoors, detect an object moving within the imaging range, and recognize patterns, characters, and the like in the object have been commercialized one after another.

  In such an image processing system, a product that recognizes characters, patterns, and the like regardless of day or night using an imaging device having a configuration of a near-infrared camera and LED illumination is generally used. In recent years, the number of systems that install a color camera together with a near-infrared camera for the purpose of displaying and storing images during recognition has begun to increase.

  However, when two cameras are installed outdoors, it is inevitable that the secrecy of the imaging device is reduced due to an increase in the surface area of the imaging device. For this reason, a system that uses a single camera to recognize characters and patterns and store images using a day / night camera capable of switching between a color / near-infrared camera is being studied.

  In the case of outdoor installation, it is a function required of an image processing system to maintain the recognition performance of characters, patterns, etc. in response to changes in sunshine conditions due to the season, weather, and the like. In order to realize this, it is necessary to capture characters and patterns to be recognized with brightness suitable for recognition processing. Therefore, the image processing system has a function of automatically adjusting the iris (aperture) of the camera as needed, and performs control to keep the recognition target at an appropriate brightness.

A technique for controlling an iris (aperture) of a camera in accordance with the brightness of an image is known (for example, see Patent Document 1). In addition, a camera internal processing technique that maintains the gradation of the entire image corresponding to changes in external light is known (see, for example, Patent Document 2).
JP-A-6-46325 Japanese Patent Laid-Open No. 7-23287

  Control that keeps the recognition target at an appropriate brightness is intended to maintain recognition performance. Depending on the outdoor environmental conditions, even if the surrounding area of the recognition target has an appropriate brightness, people can see the entire image. May result in inappropriate brightness.

  For example, it is assumed that strong direct sunlight is shining around the recognition target. In this case, the camera iris is closed to darken the image so that the brightness around the recognition target is appropriate. With this control, the surrounding area of the recognition target that is exposed to direct sunlight has an appropriate brightness, but the other parts in the imaging range may not be exposed to direct sunlight, and that part is quite It will be dark. Thereby, since the recognition target is imaged with appropriate brightness, the recognition performance can be maintained, but the visibility of the entire image is too dark.

  On the other hand, if shadows of buildings, utility poles, surrounding trees, etc. are hit around the recognition target, similarly, if the periphery of the recognition target is controlled to an appropriate brightness, the entire image will be too easy to see. .

  In other words, in order to maintain recognition performance, when control is performed so that the brightness around characters and patterns is appropriate, the overall image of the moving object has a certain level of brightness around the recognition target and in the other imaging range. Under the environmental conditions in which the difference occurs, it is too bright or too dark, which is not appropriate for use as a display / accumulation purpose for visual confirmation by humans.

  Therefore, in order to achieve both the maintenance of the recognition performance and the storage of the image using one camera, it is necessary to appropriately control the brightness of both the character and the pattern periphery and the entire captured image.

  SUMMARY OF THE INVENTION An object of the present invention is to solve the above-described problems and to provide a technique that can maintain both the recognition performance of characters and patterns from a captured image and the visibility of the entire captured image.

An image analysis device that analyzes an image captured by the imaging device includes an image capturing unit that captures the captured image from the imaging device into a memory, a specific pattern that is a recognition target by analyzing the captured image, and a recognition result. A recognition target recognition unit that outputs, a recognition target brightness correction unit that determines control for adjusting the amount of light captured by the imaging device so that the brightness value of a predetermined region including the recognition target is a value within a predetermined range; , A camera control unit that outputs a control signal for adjusting the amount of light to the imaging device, a moving object detection unit that detects a moving object from the captured image, and calculates the overall brightness value of the captured image in which no moving object was detected The captured image brightness value calculation unit and the brightness of the captured image captured in the memory are corrected so that the overall brightness value of the captured image is within a predetermined range, and the corrected captured image is output. All captured images And a visibility correction unit.

  The processing by the above image analysis apparatus can also be realized by hardware such as CPU and memory provided in the computer and a software program.

  The amount of light captured by the camera of the imaging device is automatically adjusted so that the brightness of the surrounding area of the recognition target in the captured image is a sufficient range suitable for recognition processing, so that the brightness is easy to see for people. By automatically correcting the brightness of the captured image captured in the memory, it is possible to obtain a highly accurate recognition result and an image whose brightness is corrected so that it can be easily seen by a person. That is, it is possible to maintain the character / pattern recognition performance and display / accumulate a captured image with appropriate brightness without being affected by the weather, sunshine conditions, and the like.

  In addition, since a single imaging device can be used for both the character / pattern recognition function and the image display / storage function, the cost of the device / system can be reduced.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  In this embodiment, a moving object is detected from images continuously captured using a single camera, and characters, patterns, and the like that are recognition targets are recognized in the image from which the moving object is detected. The camera iris (aperture) is automatically controlled so that the brightness of the peripheral position of the recognition target is within a certain range, and at the same time, the brightness that makes it easy for humans to visually check the brightness of the entire captured image of moving objects. An example of automatic correction will be described.

  FIG. 1 is a diagram illustrating a configuration example of an image analysis apparatus according to the present embodiment.

  The image analysis apparatus 10 captures an image captured by the imaging apparatus 20 (hereinafter referred to as a captured image) into a memory, analyzes the captured image, and recognizes a specific pattern (character, pattern, etc.) that is a recognition target. The automatic iris control of the imaging device 20 according to the brightness of the peripheral area to be recognized and the automatic brightness correction of the entire captured image taken into the memory are performed.

  The image analysis apparatus 10 includes an image capturing unit 11, a moving object detection unit 12, a character / pattern recognition unit 13, a recognition target brightness correction unit 14, a camera control unit 15, a captured image brightness value calculation unit 16, and the entire captured image. A legibility correction unit 17 is provided.

  As the imaging device 20, a day / night camera capable of shooting day and night with one camera is used. The imaging device 20 operates as a color camera during the daytime, and operates as a near infrared camera + LED illumination at night. Although not shown, the imaging device 20 has a mechanism for adjusting and controlling the amount of light (exposure) taken into the camera, such as a mechanism for controlling the iris (aperture) of the camera and a mechanism for controlling the shutter speed. Is provided.

  In the image analysis device 10, the image capturing unit 11 captures captured images continuously transferred from the imaging device 20 into the image buffer 18. The image buffer 18 has, for example, a ring buffer configuration with a plurality of surfaces, and is designed to have a number of surfaces that can hold all images from the same moving object appearing in the image to disappearing.

  The moving object detection unit 12 checks whether there is a moving object in the captured image captured by the image capturing unit 11.

  As a method for determining the presence / absence of a moving object, for example, a method of comparing consecutive images and determining that a moving object has been detected when a predetermined change or more is detected for a predetermined number of times, If the average brightness value of the image determined to have no moving object is determined as the reference brightness value, the average brightness value of the captured image is compared with the reference brightness value, and a change greater than a predetermined value is detected There are various methods such as a method of determining that a moving object has been detected.

  When it is determined that a character or pattern that is a recognition target exists in the captured image in which the moving object is detected, the character / pattern recognition unit 13 performs recognition processing on the character or pattern that is the recognition target. . Information such as the recognition result obtained by the recognition process and the position in the image to be recognized is output. When it is desired to recognize a character or pattern that is a recognition target from a moving object, the area to be recognized may be narrowed down to an area where the moving object is captured in the captured image.

  As a method of authenticating a specific pattern such as a character or pattern that is a recognition target, for example, a template of a character or pattern that is a recognition target is prepared, and matching processing in the captured image is performed using the template. There are methods.

  In addition, when the rough character, brightness, size, etc. of the character and pattern to be recognized are known, the recognition processing area in the captured image is narrowed down by using the information, and the narrowed recognition processing area. There is also a method for performing a matching process using a template.

  For example, in recent shopping malls and the like, there is a system for recognizing a vehicle number at the entrance of a parking lot in order to identify a vehicle whose parking fee has been cleared in advance. The license plate of the vehicle is composed of a combination of specific colors such as white and green, yellow and black, so specific images such as white and green, yellow and black, etc. are taken from the image of the vehicle taken at the entrance of the parking lot. If the recognition process area is narrowed down to the color combination area and the number character recognition process is performed, the necessary recognition process can be performed without waste.

  The recognition target brightness correction unit 14 adjusts the brightness of a captured image captured by the imaging device 20 so that the brightness of a predetermined recognition target peripheral region is within a predetermined range of sufficient brightness suitable for recognition processing. The control for adjusting the height, that is, the control for adjusting the amount of light (exposure) taken into the camera of the imaging device 20 is determined.

  The predetermined recognition target peripheral region indicates a predetermined region including a recognition target region in the captured image. For example, the minimum rectangular area including the recognition target may be obtained from the position in the image of the recognition target obtained by the recognition processing and used as the recognition target peripheral area, or the recognition target area + the surrounding area of several pixels may be recognized. It may be a peripheral area. The recognition processing area narrowed down in the captured image may be a recognition target peripheral area.

  There are various possible values indicating the brightness of the recognition target peripheral area. In this embodiment, the average brightness value of each recognition target peripheral area (the luminance value for each pixel) is used as the brightness value of the recognition target peripheral area. Are obtained by adding all the numbers and dividing by the number of pixels. That is, the recognition target brightness correction unit 14 makes the average brightness value of the recognition target peripheral region to be a value within the predetermined range when the average luminance value of the recognition target peripheral region is not a value within the predetermined range. Control for adjusting the amount of light taken into the imaging device 20 is determined.

  The camera control unit 15 uses the recognition target brightness correction unit 14 to determine the brightness value (in this case, the average luminance value) of the recognition target peripheral region within the predetermined range. A control signal for adjusting the amount is output to the imaging device 20.

  In the present embodiment, iris control of the imaging device 20 is performed as control for adjusting the amount of light taken into the imaging device 20. Note that as control for adjusting the amount of light taken into the imaging device 20, control of the shutter speed of the imaging device 20 or control combining iris and shutter speed may be performed. For example, when it is assumed that the moving speed of the moving object is high, blurring of the subject occurs when the shutter speed is slow. Therefore, it is desirable to adjust the amount of light taken into the imaging device 20 by iris control.

  For example, it is assumed that the iris adjustment of the imaging device 20 is set to several levels in advance.

  In the image analysis apparatus 10, the recognition target brightness correction unit 14 obtains an average luminance value of the recognition target peripheral area. When the obtained average luminance value of the recognition target peripheral region is not within a predetermined brightness range, a difference between the average luminance value of the recognition target peripheral region and a predetermined reference value is calculated. Here, it is assumed that the predetermined reference value is an intermediate value of a predetermined range with sufficient brightness suitable for recognition processing.

  For example, in the image analysis device 10, it is assumed that correspondence information between the calculated difference and the number of iris adjustment levels of the imaging device 20 is held in advance. The recognition target brightness correction unit 14 obtains the amount of light taken into the imaging device 20 by obtaining the number of iris adjustment levels such as how many levels the iris of the imaging device 20 is opened and how many levels are reduced from the calculated difference. Determine the control to be adjusted.

  The camera control unit 15 is a control signal for adjusting the amount of light taken into the imaging device 20 determined by the recognition target brightness correction unit 14, and here, iris control such as how many levels the iris of the imaging device 20 is opened and how much is reduced. The signal is output to the imaging device 20.

  In the imaging device 20, when an iris control signal is received from the image analysis device 10, the mechanism that actually controls the iris of the camera controls the iris adjustment according to the instruction of the iris control signal.

  Thus, whenever the recognition process is performed by the character / pattern recognition unit 13, the brightness of the captured image obtained from the imaging apparatus 20 by the recognition target brightness correction unit 14 and the camera control unit 15 is determined in the recognition process. It is automatically adjusted to a suitable brightness.

  The captured image brightness value calculation unit 16 obtains the overall brightness of the captured image in which no moving object has been detected. Hereinafter, a captured image in which no moving object is detected is referred to as a background image. Although various values indicating the brightness of the background image can be considered, in this embodiment, the average luminance value of the background image is obtained as the brightness value of the background image. It is assumed that the average luminance value of the background image is calculated at a predetermined interval, and is notified to the entire captured image viewability correcting unit 17 as needed.

  Here, although the captured image brightness value calculation unit 16 calculates the brightness of the entire captured image in which no moving object is detected, the brightness of the entire captured image is determined regardless of the presence or absence of the moving object. It may be. By obtaining the brightness of the entire captured image in which the moving object has not been detected, the brightness of the entire captured image according to the location and the time without including noise due to the moving object can be obtained.

  The entire captured image viewability correcting unit 17 captures the brightness value of the background image notified at a predetermined interval into the image buffer 18 so that the brightness value is within a predetermined range suitable for human viewing. The brightness of the captured image is corrected, and the corrected captured image is output.

  For example, when the average brightness value of the latest background image calculated by the captured image brightness value calculation unit 16 is not within a predetermined brightness range, the overall captured image viewability correction unit 17 determines the average brightness value of the background image. Is calculated with respect to a predetermined reference value. Here, it is assumed that the predetermined reference value is an intermediate value in a predetermined range with sufficient brightness suitable for human viewing.

  The entire captured image viewability correcting unit 17 calculates the difference calculated for each pixel of the captured image, for example, so that the captured image becomes bright when the entire captured image is too dark, and the captured image becomes dark when the entire captured image is too bright. The brightness of the entire captured image captured in the image buffer 18 is corrected by adjusting the luminance according to the amount. A brightness-corrected captured image (overall image) is output.

  In this way, the captured image captured by the whole image captured visibility correction unit 17 is automatically adjusted to an image having brightness suitable for human viewing and output so that the brightness is suitable for recognition processing. Is done.

  Note that the entire captured image viewability correction unit 17 may correct the overall brightness of the captured image captured in the image buffer 18, or a portion other than a region where a character or a pattern is recognized. The brightness may be corrected. If the brightness of the portion excluding the area where the character or pattern is recognized is corrected, an image that is easy to see as a whole image and easy to see the character or pattern portion can be obtained.

  The recognition result obtained by the image analysis apparatus 10 according to the present embodiment and the captured image whose brightness has been corrected are output to a network, a display / storage medium, or the like. As a result, a highly accurate recognition result of characters, patterns, etc. and an entire image with appropriate brightness can be obtained.

  FIG. 2 is an image analysis processing flowchart according to the present embodiment. The process illustrated in the example of FIG. 2 is an image analysis process that is executed on each of the captured images continuously transmitted from the imaging device 20.

  In the image analysis device 10, the image capturing unit 11 captures the captured image sent from the imaging device 20 into the image buffer 18 (step S10). The moving object detection unit 12 confirms whether a moving object has been detected from the captured image captured in the image buffer 18 (step S11).

  When a moving object is not detected (NO in step S11), if a predetermined timing such as an interval of 1 minute (YES in step S12), the captured image brightness value calculation unit 16 detects the moving object. The overall brightness of the captured image that has not existed, here, the average luminance value is calculated (step S13).

  If a moving object is detected (YES in step S11), the character / pattern recognition unit 13 performs a recognition process for a specific pattern (character, pattern, etc.) to be recognized (step S14).

  When the recognition target is recognized (YES in step S15), the recognition target brightness correction unit 14 calculates the brightness of the recognition target peripheral area, here the average luminance value (step S16), and the recognition target peripheral It is confirmed whether the brightness of the area is within a predetermined range (step S17).

  If the brightness of the recognition target peripheral area is not within a predetermined range (NO in step S17), the recognition target brightness correction unit 14 sets the brightness of the recognition target peripheral area to a brightness within the predetermined range. In this way, control for adjusting the amount of light taken in by the imaging device 20, that is, control for iris adjustment of the imaging device 20 is determined (step S18). The camera control unit 15 outputs an iris control signal to the imaging device 20 (step S19).

  If the brightness of the recognition target peripheral area is within a predetermined range (YES in step S17), the entire captured image viewability correction unit 17 determines that the brightness of the latest captured image obtained in step S13 is It is confirmed whether the brightness is within a predetermined range (step S20). If the brightness of the entire captured image is not within the predetermined range (NO in step S20), the brightness of the entire captured image is predetermined. The overall brightness of the captured image captured in the image buffer 18 is corrected so that the brightness is within the range (step S21).

  The image analysis apparatus 10 outputs the recognition result and the captured image to a network, a display / storage medium, or the like (step S22). The captured image output here is a captured image from which a recognition target has been detected, and is corrected to a brightness that is easy for humans to see.

  By the processing of the image analysis apparatus 10 as described above, the recognition result of characters / patterns and the like recognized from the moving object in the captured image with high accuracy and the brightness of the moving object including the recognition target corrected to the human-readable brightness are included. A captured image is obtained.

  FIG. 3 is a process flowchart illustrating a more specific example of recognition target brightness correction, camera control, and overall captured image visibility correction by the image analysis apparatus of the present embodiment.

First, the character / pattern recognition unit 13 executes recognition processing of characters and patterns that are recognition targets on the captured image extracted by detecting the moving object, and obtains the position in the captured image of the recognition target. The brightness (average luminance value N ₁ ) of the circumscribed rectangular area (recognition target peripheral area Z ₁ ) around the recognition target is calculated by the recognition target brightness correction unit 14 based on the obtained position of the recognition target. (Step S30).

The image analysis apparatus 10 sets the iris value of the camera so that the calculated value N ₁ falls within a certain range suitable for recognition processing. Here, it is assumed that a certain range suitable for recognition processing is represented by N ₀ ± Th _N using an intermediate value of the range as a reference value N ₀ and a threshold value Th _N.

That is, the recognition target brightness correction unit 14 determines that the calculated average luminance value N _{1 of} the recognition target peripheral region Z ₁ is not a value within a certain range N ₀ ± Th _N suitable for recognition processing (step S31). NO), the brightness of the recognition target peripheral area Z ₁ is suitable for the recognition process according to the difference value (N ₁ −N ₀ ) between the average luminance value N _{1 of} the recognition target peripheral area Z ₁ and the reference value N _0. Then, the iris adjustment control of the imaging device 20 is determined so as to be within the predetermined range (step S32), and the camera control unit 15 outputs the iris control signal to the imaging device 20 (step S33).

  Up to this point, the brightness control processing is performed on the image captured by the imaging device 20 in order to maintain the accuracy of the recognition processing. Next, the image analysis device 10 determines the appropriateness of the brightness of the entire captured image captured from the imaging device 20, and corrects the entire captured image so that the brightness becomes easy to see.

  First, the image analysis apparatus 10 checks whether the current time is within a predetermined time (step S34). Here, it is confirmed whether it is during the day. The day / night camera operates as a color camera during the daytime and operates as a near-infrared camera at nighttime.

  Daytime is a time zone in which brightness differences occur in the captured image due to direct sunlight and shadows from surrounding objects, so there is a difference in brightness between the area where the recognition target is shown and other areas. May occur. For example, 5:00 to 18:00 is set as the predetermined time, and an image captured during that time period is set as an object of the whole captured image visibility correction of the present embodiment. The predetermined time can be changed according to the season, or can be automatically determined based on ambient brightness.

  At night, it is common to use a dedicated lighting device (LED lighting, etc.) in order to ensure appropriate brightness for recognition processing. In that case, there is not much difference in the brightness in the image. Does not occur. For this reason, in the present embodiment, the image captured at night is not subject to correction of the visibility of the entire captured image. If the current time is not within the predetermined time (NO in step S34), the image analysis apparatus 10 transmits the captured image as it is to the storage medium without correcting the brightness of the captured image (step S39).

If the current time is within a predetermined time (YES in step S34), the captured image overall visibility correction unit 17 obtains the latest value of the average luminance value A ₁ of the background image to be notified at regular intervals (step S35 ), And the reference value A ₀ of the brightness of the background image that is formulated in a timely manner according to the system, environmental conditions, etc. (step S36).

Here, the reference value A ₀ of the brightness of the background image is a reference suitable for an individual environment by weighted averaging of the average luminance value of the background image during the time period when no brightness correction is performed. Can be automatically adjusted to the value. The reference value A ₀ for the brightness of the background image is adjusted, for example, according to the following procedures a) to d).

a) Calculate the average luminance value of the entire captured image every minute b) Average the calculated average luminance value of the entire captured image in units of 10 minutes, and perform statistical management along with the presence or absence of correction of the captured image in that time zone a weighted average value between the average value a _V and the calculated reference value a ₀ at d) the moment of calculating the to keep c) the average value a _V correction average luminance value of the unnecessary only when the background image of the captured image thus the next day of the reference values a _0, can be adjusted at each location, the reference value a ₀ which is suitable for time.

Returning to the description of FIG. 3, when the average luminance value A ₁ of the background image is within a certain range (± threshold Th _A ) around the reference value A ₀ (YES in step S36), the entire captured image is easily viewed. The unit 17 determines that the brightness of the entire captured image is appropriate, and the image analysis apparatus 10 transmits the captured image as it is to the storage medium without correcting the brightness of the captured image (step S39).

When the average luminance value A ₁ of the background image is not within a certain range (± threshold Th _A ) around the reference value A ₀ (NO in step S36), the entire captured image viewability correcting unit 17 It is determined that the brightness is inappropriate, and the brightness correction reference value A ₂ (= A ₁ −A ₀ ) is calculated (step S37).

The brightness correction reference value A ₂ is a difference between the average brightness value A ₁ of the background image and the brightness reference value A ₀ of the background image. In order to suppress the correction amount within a certain range, the value of | A ₂ | _Is less than or equal to the maximum value Am. That is, if | A ₂ |> A _m , | A ₂ | = A _m . Note that the maximum value _Am is determined in a timely manner according to the system.

When the average luminance value A ₁ of the background image is far from the reference value A _0, there is a possibility that some noise is included, so the upper limit of the luminance correction reference value A ₂ (the lower limit in the case of minus) ) Is set, the brightness correction amount of the entire captured image is suppressed within a certain range.

The entire captured image visibility correction unit 17 corrects the luminance value of the entire recognized image in accordance with the calculated luminance correction reference value A ₂ (step S38). Here, the recognized image is a captured image in which characters / patterns to be recognized are recognized. The image analysis apparatus 10 transmits the captured image whose brightness has been corrected to a storage medium or the like (step S39).

When the luminance correction reference value A ₂ is equal to or greater than a certain value, gamma correction is performed before the luminance correction, and brightness correction that is easier to see can be performed by suppressing portions that are too bright or too dark. The intensity of the gamma correction, providing several stages of the table according to the range of luminance correction reference value A _2.

  It should be noted that various corrections can be made for the brightness correction of the captured image captured by the image analysis apparatus 10 depending on the required function of the system. An example is shown below.

  a) The brightness of the entire captured image is corrected. This is the brightness correction of the captured image in the example shown in FIG.

  b) The brightness only around the area of the captured image that is too bright or too dark is corrected. For example, when it is necessary to confirm details of the entire captured image, such brightness correction is effective.

  c) Brightness correction is performed only on a specific position in the captured image based on the position of the character / pattern to be recognized from the shape characteristics of the moving object including the recognition target. For example, such brightness correction is effective when it is necessary to confirm details of a specific part of a moving object.

  FIG. 4 is a process flowchart illustrating a specific example in the case of correcting the brightness of a region other than the periphery of the recognition target of the captured image when the entire captured image is easily viewed by the image analysis apparatus according to the present embodiment.

  In FIG. 4, the processing from step S50 to step S57 is the same as the processing from step S30 to step S37 in FIG.

The entire captured image visibility correction unit 17 calculates a circumscribed rectangular region (recognition target peripheral region Z ₂ ) around the recognition target from the coordinates on the captured image of the character / pattern that is the recognition target (step S58). The recognition target peripheral region Z _{2 that} is the _second predetermined region including the recognition target is the recognition target peripheral region Z ₁ that is the first predetermined region including the recognition target, for which the average luminance value N ₁ is obtained in step S50. May be the same region or different regions.

Captured image overall visibility correction unit 17, the recognition image, selects a pixel to be processed (step S59), if the pixel to be processed which is selected recognition target peripheral region Z ₂ inside the pixel (step S60 in NO), according to the luminance correction reference value a ₂ obtained in step S57, the correcting luminance values of the pixels to be the processing target (step S61). The entire captured image visibility correction unit 17 selects a pixel to be processed next (step S59), and performs the processing of step S59 to step S61 for all the pixels in the recognized image (step S62).

The image analysis device 10 transmits the captured image whose brightness has been corrected to a storage medium or the like (step S63). Here, the image whose brightness is corrected is a captured image in which the brightness of the area other than the recognition target peripheral area Z ₂ is corrected. By such processing by the image analysis apparatus 10, it is possible to obtain a captured image that maintains the accuracy of recognition of characters and patterns to be recognized, and that allows easy recognition of the characters and patterns that are to be recognized and the entire image.

  Even when imaging is performed using one camera by the image analysis apparatus 10 according to the present embodiment, character / pattern recognition performance is maintained by appropriate camera iris control, and brightness suitable for display / accumulation. The corrected image can be obtained.

  The processing by the image analysis apparatus 10 described above can be realized by hardware such as CPU and memory provided in the computer and a software program, and the program can be recorded on a computer-readable recording medium through a network. It is also possible to do.

  Although the present embodiment has been described above, the present invention can naturally be modified in various ways within the scope of the gist thereof.

  For example, in the image analysis apparatus 10 according to the present embodiment, a moving object is detected from a captured image captured in a memory, and a specific pattern that is a recognition target is recognized from the captured image from which the moving object is detected. However, it does not detect a moving object, recognizes a specific pattern that is a recognition target from a captured image captured in a memory regardless of the presence or absence of the moving object, and based on the brightness of a predetermined area including the recognition target. Thus, the amount of light taken into the imaging device 20 may be controlled.

  If a specific pattern that is a recognition target is recognized from a captured image in which a moving object is detected, and the amount of light taken into the imaging device 20 is controlled based on the brightness of a predetermined area including the recognition target. For example, it is possible to maintain the recognition performance of a specific pattern which is a recognition target on a moving object. Regardless of the presence or absence of a moving object, a specific pattern that is a recognition target is recognized from the captured image captured in the memory, and the amount of light captured by the imaging device 20 is controlled based on the brightness of a predetermined area including the recognition target. By doing so, for example, it is possible to maintain the recognition performance of a specific pattern that is a recognition target on the background of the captured image.

It is a figure which shows the structural example of the image analysis apparatus by this Embodiment. It is an image analysis process flowchart by this Embodiment. It is a process flowchart which shows the more specific example of recognition object brightness correction by the image analysis apparatus of this Embodiment, camera control, and the whole captured image visibility correction | amendment. It is a process flowchart which shows the specific example in the case of performing brightness correction of area | regions other than the periphery of the recognition target of a captured image at the time of the whole image captured image visibility correction | amendment by the image analysis apparatus of this Embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Image analyzer 11 Image capture part 12 Moving object detection part 13 Character / pattern recognition part 14 Recognition target brightness correction part 15 Camera control part 16 Captured image brightness value calculation part 17 Whole captured image viewability correction part 18 Image buffer 20 Imaging device

Claims (5)

An image analysis device for analyzing an image captured by an imaging device,
An image capturing unit that captures a captured image from the imaging device into a memory;
A recognition target recognition unit that analyzes the captured image, recognizes a specific pattern that is a recognition target, and outputs a recognition result;
A recognition target brightness correction unit that determines control for adjusting the amount of light taken in by the imaging device so that the brightness value of the predetermined area including the recognition target is a value within a predetermined range;
A camera control unit for outputting a control signal for adjusting the light amount to the imaging device;
A moving object detection unit for detecting a moving object from the captured image;
A captured image brightness value calculating unit that calculates the overall brightness value of the captured image in which no moving object is detected ;
Correcting the brightness of the captured image captured in the memory so that the overall brightness value of the captured image is within a predetermined range, and outputting the corrected captured image An image analysis apparatus comprising: a unit. The image analysis according to claim 1, wherein the whole captured image viewability correction unit corrects brightness of a portion of the captured image captured in the memory, excluding a predetermined region including the recognition target. apparatus. Before Symbol recognition target recognition unit analyzes the captured image moving objects are detected, the image analysis apparatus according to claim 1 or 2, characterized in that recognize specific pattern which is the recognition target. The image analysis apparatus according to any one of claims 1 to 3 , wherein the control for adjusting the amount of light is an iris control of the imaging apparatus. A program executed by a computer of an image analysis device that analyzes an image captured by an imaging device,
Said computer,
An image capturing unit that captures a captured image from the imaging device into a memory;
A recognition target recognition unit that analyzes the captured image, recognizes a specific pattern that is a recognition target, and outputs a recognition result;
A recognition target brightness correction unit that determines control for adjusting the amount of light taken in by the imaging device so that the brightness value of the predetermined area including the recognition target is a value within a predetermined range;
A camera control unit for outputting a control signal for adjusting the light amount to the imaging device;
A moving object detection unit for detecting a moving object from the captured image;
A captured image brightness value calculating unit that calculates the overall brightness value of the captured image in which no moving object is detected ;
Correcting the brightness of the captured image captured in the memory so that the overall brightness value of the captured image is within a predetermined range, and outputting the corrected captured image Image analysis program to function as a part.

Images