JP2003259217A - Infrared camera - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an infrared camera capable of correcting output irregularities of each pixel while always imaging an object to be imaged. <P>SOLUTION: A correction candidate specifying part 11 of a processor 6 regards a pixel, the pixel value of which is equal to or less than a prescribed threshold Vp as a pixel obtained by imaging a background in an image to specify the pixel as a correction candidate among each of pixels of an infrared sensor 5. A difference calculating part 12 sets for a pixel, the pixel value of which is different from a reference value Vr, its difference as a correction coefficient among pixels of the correction candidate. A correction processing part 13 corrects the pixel value of the pixel with the correction coefficient set therein by using the correction coefficient and outputs the corrected value as an output value. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an infrared camera that captures and outputs an infrared image using an infrared sensor, and more particularly to an infrared camera having a function of correcting a pixel value for each pixel of the infrared sensor. .

[0002]

2. Description of the Related Art In recent years, along with the intelligentization of vehicles, attempts are being made to mount an infrared camera on the vehicle and monitor the surroundings of the vehicle using an image captured by the infrared camera. The image captured by the infrared camera is output as an image that is color-coded according to the temperature of the imaging target, so if there is an object having a temperature such as a human body around the vehicle and the object is captured by the infrared camera, for example. With reference to the image output from the infrared camera, the presence of an object can be properly recognized.

By the way, the infrared camera is designed to convert the incident infrared rays incident on each pixel of the infrared sensor formed by using a semiconductor into a pixel value (voltage value) for each pixel and output it. When the ambient temperature of the infrared sensor changes, the output from each pixel may fluctuate due to the temperature characteristics of the infrared sensor, and an appropriate image may not be obtained.

Therefore, as a method of correcting the output variation due to the temperature change of such an infrared camera, for example, a technique disclosed in Japanese Patent Laid-Open No. 9-325073 has been proposed. The infrared camera described in Japanese Patent Application Laid-Open No. 9-325073 is provided with a shutter for exclusive use of correction for blocking incident infrared light in front of the infrared sensor, and when the temperature change around the infrared sensor becomes a predetermined value or more,
Close the shutter dedicated to correction to capture a thermally uniform surface,
Based on the result, the correction coefficient is set for each pixel so that the output values of each pixel are equal.

In the infrared camera as described above, the pixel value for each pixel is corrected by the correction coefficient set for each pixel and output, so that an appropriate image faithfully reflecting the temperature of the object to be imaged is obtained. Will be obtained.

[0006]

However, in the infrared camera as described above, when correcting the output variation due to the temperature change, the shutter dedicated to the correction is closed to capture a thermally uniform surface. Therefore, there is a problem in that during that time, it is not possible to capture an image of an object to be originally captured. In particular, an infrared camera tends to have a large temperature change around the infrared sensor immediately after startup, and it is necessary to frequently perform correction processing immediately after startup, so that it takes a long time to image an object to be originally imaged.

When an infrared camera is used to monitor the surroundings of a vehicle, especially when the vehicle starts traveling,
Since it is required to properly recognize whether or not there is an object around the vehicle based on the image from the infrared camera, it is strongly desired that the object to be imaged can always be imaged.

The present invention was conceived in view of the above circumstances, and it is an object of the present invention to provide an infrared camera capable of correcting variations in output for each pixel while always capturing an image of an object to be imaged. Has an aim.

[0009]

According to a first aspect of the present invention, there is provided an infrared sensor having a plurality of pixels, each pixel outputting a pixel value corresponding to an incident infrared ray, and the infrared sensor outputting the pixel value. Comparing each pixel value with a predetermined threshold value, the pixel value is specified as a correction candidate pixel below the threshold value, for the pixels of the specified correction candidate pixel value is different from the predetermined reference value, For the correction coefficient setting means for setting the difference between the pixel value and the reference value as the correction coefficient, and for the pixel for which the correction coefficient is not set by the correction coefficient setting means, the pixel value output by the infrared sensor is used as the output value. For the pixel which is output and the correction coefficient is set by the correction coefficient setting means, the pixel value output by the infrared sensor is corrected by using the correction coefficient and is output as an output value. An infrared camera, characterized in that it comprises an image output means.

In the infrared camera according to the first aspect of the present invention, when the infrared ray from the object to be imaged is incident on each pixel of the infrared sensor, the infrared sensor outputs a pixel value corresponding to the incident infrared ray of each pixel. The pixel value for each pixel output from the infrared sensor is supplied to the correction coefficient setting means and compared with a predetermined threshold value by the correction coefficient setting means. And
Pixels of the infrared sensor that output pixel values below the threshold are specified as correction candidates. Here, the threshold value to be compared and contrasted with the pixel value is set to an optimum value capable of separating the object and the background. A pixel that outputs a pixel value equal to or less than the threshold is regarded as an image of the background, and a pixel that is considered to have imaged the background is set as a correction candidate. That is, in this infrared camera, the correction process is performed only when necessary on the pixels that are considered to have captured the background, and the correction process is not performed on the pixels that are considered to have captured the object. I have to.

When the correction candidate pixel is specified, the correction coefficient setting means compares the pixel value of the correction candidate pixel with a predetermined reference value. Here, the reference value is set to a value equal to the pixel value obtained when a uniform background is imaged. Then, of the correction candidate pixels, the difference between the pixel value and the reference value is set as a correction coefficient for the pixel whose pixel value is different from the predetermined reference value. That is, the correction candidate pixel is a pixel regarded as an image of a uniform background, and the pixel value should originally be equal to the reference value. Therefore, a pixel whose pixel value is different from the reference value can be regarded as a pixel that needs correction, and the correction coefficient setting means
For such a pixel, the difference between the pixel value and the reference value is set as a correction coefficient.

When the correction coefficient is set for the pixel that needs to be corrected, the image output means then outputs the output value for each pixel for displaying the image. At this time, the image output means, for the pixel for which the correction coefficient is not set,
The pixel value from the infrared sensor is output as it is as an output value, and for the pixel for which the correction coefficient is set, the pixel value from the infrared sensor corrected using the correction coefficient is output as the output value. As a result, the output variation for each pixel due to the temperature change or the like is corrected, and an appropriate image display can be performed.

The invention described in claim 2 is the same as claim 1
In the infrared camera described in (3), the correction coefficient setting unit performs the correction coefficient setting process for each image output cycle.

In the infrared camera according to the second aspect, the correction coefficient is set for each pixel for which the pixel value needs to be corrected, and the pixel value corrected using the latest correction coefficient is always set. It will be output as an output value.

The invention according to claim 3 is the same as claim 1
In the infrared camera described in (3), the correction coefficient setting means performs the correction coefficient setting process when the amount of change in the output image becomes large.

In the infrared camera according to the third aspect, when the amount of change in the output image becomes large, a correction coefficient is set for a pixel whose pixel value needs to be corrected, and the pixel corrected by this correction coefficient is set. The value will be output as the output value.

The invention according to claim 4 is the same as claim 1.
In the infrared camera described in (3), the correction coefficient setting means performs the correction coefficient setting process when the amount of change in the ambient temperature of the infrared sensor becomes large.

In the infrared camera according to the fourth aspect, when the change amount of the ambient temperature of the infrared sensor becomes large,
A correction coefficient is set for a pixel whose pixel value needs to be corrected, and the pixel value corrected by this correction coefficient is output as an output value.

The invention described in claim 5 is the same as claim 1.
5. In the infrared camera according to any one of 4 to 4, the image output unit corrects by the correction coefficient setting unit by a calculation process including a simple average, a weighted average, or a moving average with a correction factor set in the past. It is characterized in that the pixel value of the pixel for which the coefficient is set is corrected.

In the infrared camera according to the present invention, the pixel value of the pixel whose pixel value needs to be corrected is corrected in consideration of the correction coefficient previously set for this pixel and is output as the output value. Will be done.

The invention according to claim 6 is the same as claim 1.
4. In the infrared camera according to any one of 4 to 4, the image output means is configured to perform a calculation process including a simple average, a weighted average, or a moving average of a preset initial correction coefficient and a correction coefficient set in the past. The pixel value of the pixel for which the correction coefficient is set by the correction coefficient setting means is corrected.

In the infrared camera according to the present invention, the pixel value of the pixel whose pixel value needs to be corrected takes into consideration the initial correction coefficient preset for this pixel and the correction coefficient set in the past. Will be corrected and output as an output value.

[0023]

According to the infrared camera of the first aspect,
Of the pixels of the infrared sensor, only for the pixels that are considered to be capturing the background, set a correction coefficient when correction is necessary, and use this correction coefficient to correct the pixel value and output the value. Therefore, it is possible to correct an output variation for each pixel while always capturing an image of an imaging target and display an appropriate image without interrupting the imaging when setting the correction coefficient.

Further, according to the infrared camera of the second aspect, since the correction coefficient for the pixel whose pixel value needs to be corrected is set for each image output cycle, this pixel is always updated with the latest correction coefficient. Can be used for correction, and an appropriate image can always be displayed.

Further, according to the infrared camera of the third aspect, when the amount of change in the output image becomes large, the correction coefficient is set for the pixel whose pixel value needs to be corrected. Efficient correction processing can be performed while reducing the processing load.

Further, according to the infrared camera of the fourth aspect, the correction coefficient is set for the pixel whose pixel value needs to be corrected when the change amount of the ambient temperature of the infrared sensor becomes large. Therefore, efficient correction processing can be performed while reducing the processing load.

According to the infrared camera of the fifth aspect, the correction of the pixel value includes a calculation process including a simple average, a weighted average, or a moving average with a correction coefficient previously set for the pixel. Therefore, even if there is an output fluctuation due to a sudden noise mixture or the like, it is possible to perform an appropriate correction process without being influenced by such a sudden factor.

Further, according to the infrared camera of the sixth aspect, the pixel value is corrected by a simple average or a weighted average of the initial correction coefficient preset for the pixel and the correction coefficient set in the past, or Since it is performed by arithmetic processing including moving average, even if there is an output change due to sudden noise mixing, etc., appropriate correction processing is not affected by such sudden factors. By adding the initial correction coefficient, it is possible to appropriately correct the fluctuation component peculiar to the pixel.

[0029]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings. Here, an example will be described in which the present invention is applied to a vehicle surroundings monitoring device for displaying the image around the vehicle captured by an infrared camera to monitor the surroundings of the vehicle, but the present invention will be described here. However, the present invention can be effectively applied to any device that outputs an image based on the output from the infrared sensor.

FIG. 1 shows a configuration example of a vehicle surroundings monitoring device to which the present invention is applied. The vehicle surroundings monitoring device 1 shown in FIG.
Is an infrared camera 2 that captures an image of an object around the vehicle,
A display device 3 that displays an image based on the output from the infrared camera 2 is provided.

The infrared camera 2 includes an image forming optical system 4, an infrared sensor 5, a processing device 6, and a storage device 7. The imaging optical system 4 has, for example, a reflection mirror, an imaging lens, and the like,
Infrared rays emitted from the imaging target are guided to the infrared sensor 5,
The infrared sensor 5 is made to enter.

As the infrared sensor 5, for example, a thermal infrared sensor such as a thermopile infrared sensor is used. This infrared sensor 5 has a plurality of pixels,
The incident infrared rays entering each pixel are converted into heat quantity for each pixel, and the pixel value (voltage value) corresponding to the infrared ray quantity radiated from the imaging target is changed by utilizing the change in electric resistance according to the heat quantity. It is designed to output for each pixel. The pixel value output for each pixel of the infrared sensor 5 is supplied to the processing device 6.

The processing device 6 controls the operation of the infrared camera 2 as a whole, performs various arithmetic processing based on each pixel value from the infrared sensor 5, and outputs an image signal for displaying an image. Is. In particular, this processing device 6
While the image pickup operation of the infrared camera 2 is continued, the pixel of each pixel of the infrared sensor 5 that needs correction is determined and the pixel value from the pixel is corrected. Each pixel value from the infrared sensor 5 is A
After being D / D-converted and subjected to correction processing and the like by the processing device 6 as digital data, it is D / A-converted and output from the processing device 6 as an analog image signal.

The storage device 7 stores information necessary for various arithmetic processes by the processing device 6, for example, various information such as a correction coefficient set for a pixel that needs correction.

As the display device 3, for example, a liquid crystal display panel or the like is used. The display device 3 is driven based on the image signal output from the infrared camera 2, and displays the image of the surroundings of the vehicle captured by the infrared camera 2.

Here, in the vehicle surroundings monitoring apparatus 1 to which the present invention is applied, the pixel value correction processing executed by the processing unit 6 of the infrared camera 2 will be described with reference to a specific example.

The characteristics of the infrared sensor 5 of the infrared camera 2 are influenced by the ambient temperature, and the pixel value output for each pixel may change due to the temperature change. When the pixel value output from the infrared sensor 5 thus fluctuates, it is required to correct the pixel value to an appropriate value by removing the fluctuation component due to the temperature change. In order to correct such a pixel value, there is also a method of temporarily suspending image capturing, capturing a thermally uniform surface, and specifying a pixel that needs to be corrected based on each pixel value at that time. Although conceivable, it is necessary to temporarily suspend the image capturing when the pixel value is corrected. Therefore, particularly in an apparatus such as the vehicle surroundings monitoring apparatus 1 which is required to constantly monitor the surroundings of the vehicle. Adopting such a method is not desirable.

Therefore, in the vehicle surroundings monitoring device 1 to which the present invention is applied, the processing device 6 of the infrared camera 2 detects the object and the background in the captured image based on the pixel value from the infrared sensor 5. Only pixels that are considered to have captured the background are identified as correction candidates, and a correction coefficient is set for pixels that require correction among the pixels of the correction candidates that are considered to have captured the background. By performing the correction process on the pixel value using the correction coefficient, the pixel value from the pixel that needs to be corrected among the pixels of the infrared sensor 5 is corrected while continuing the image capturing operation. There is.

More specifically, as shown in the functional block diagram of FIG. 2, the processing device 6 includes a correction candidate specifying section 11, a difference calculating section 12, and a correction processing section 13. And
The correction candidate specifying unit 11 and the difference calculating unit 12 function as correction coefficient setting means, and the correction candidate specifying unit 11 and the difference calculating unit 12 set the correction coefficient for the pixel that needs correction by the infrared sensor 5. It has become so. Further, the correction processing unit 13 functions as an image output unit, and by the processing of the correction processing unit 13, the pixel value from the pixel for which the correction coefficient is set is corrected, and the corrected value is output as the output value. It has become so.

As shown in FIG. 3, the correction candidate specifying unit 11 compares the pixel value supplied to each pixel of the infrared sensor 5 with a predetermined threshold value Vp and outputs a pixel value equal to or less than the predetermined threshold value Vp. The pixels of the infrared sensor 5 that have been selected are specified as correction candidates. Here, the threshold value Vp is set to an optimum value that can separate the object and the background in the image.

Since the background portion of the image is the portion where the infrared rays emitted from the object do not enter, the pixel value of the pixel of the infrared sensor 5 which has captured the background is essentially the same value as the reference value Vr. Even if the infrared sensor 5 fluctuates due to a change in ambient temperature, the value is close to the reference value Vr. On the other hand, the infrared ray emitted from this object is incident on the portion of the image where some object is present. Therefore, the pixel value of the pixel of the infrared sensor 5 that images this object is a value distant from the reference value Vr. Becomes Therefore, by setting the threshold value Vp to a value slightly higher than the reference value Vr and comparing the pixel value supplied for each pixel of the infrared sensor 5 with the threshold value Vp, each pixel of the infrared sensor 5 images the background. It is possible to determine whether the object is one or an object is imaged. Then, the correction candidate specifying unit 11
Is the threshold value V among the pixels of the infrared sensor 5.
Pixels equal to or smaller than p, that is, pixels that are considered to have captured the background in the image are specified as correction candidates. Specifically, for example, in the example shown in FIG. 3, the pixel a, the pixel b, the pixel c, and the pixel d are each specified as a correction candidate.

The difference calculating unit 12 sets the pixel value of the pixel specified as the correction candidate by the correction candidate specifying unit 11 to the reference value V.
A pixel whose pixel value is different from the reference value Vr as compared with r is recognized as a pixel to be corrected. Then, the difference between the pixel value of the correction target pixel and the reference value Vr is calculated, and the calculated value is set as the correction coefficient of the correction target pixel. That is, as described above, the pixel specified as the correction candidate by the correction candidate specifying unit 11 is a pixel that is considered to have captured the background in the image, and therefore its output value (pixel value) is
Originally, the value is equal to the reference value Vr. Therefore, among the correction candidate pixels, those having a pixel value different from the reference value Vr are pixels requiring correction, and the pixel value and the reference value Vr
It is considered that the difference between and is the necessary correction amount. Therefore,
Among the pixels specified as correction candidates by the correction candidate specifying unit 11, the difference calculation unit 12 sets the difference between the pixel value and the reference value Vr as a correction coefficient for a pixel whose pixel value is different from the reference value Vr. I am trying to set it.

The processes of the correction candidate specifying unit 11 and the difference calculating unit 12 are performed, for example, at every image output cycle of the infrared camera 2. That is, the infrared sensor 5 outputs the pixel value of each pixel forming one image, and when the image signal for displaying the image on the display device 3 is generated based on this pixel value, the correction is performed each time. A new correction coefficient for the target pixel is set. In this case, the pixel value of the correction target pixel can always be corrected using the latest correction coefficient.

The processing of the correction candidate specifying unit 11 and the difference calculating unit 12 may be performed, for example, when the amount of change in the output image becomes large. That is, since the change in each pixel value is small while the change amount of the output image is small,
It is considered that the correction can be appropriately performed even if the correction coefficient is set for the pixel to be corrected using the correction coefficient set in the past. Therefore, the setting process of the new correction coefficient may not be performed while the change amount of the output image is small. If the correction coefficient is set when the amount of change in the output image becomes large as in this example, the processing load on the processing device 6 can be reduced. The amount of change in the output image can be determined based on the amount of change in each pixel value forming the image.

The processing of the correction candidate specifying unit 11 and the difference calculating unit 12 may be performed, for example, when the amount of change in the ambient temperature of the infrared sensor 5 becomes large.
That is, as described above, the main cause of the fluctuation in the pixel value output from each pixel of the infrared sensor 5 is due to the temperature characteristics of the infrared sensor 5, and the ambient temperature of the infrared sensor 5 greatly changes. It is considered that the pixel value from each pixel of the infrared sensor 5 does not change so much while it is not performing. Therefore, while the amount of change in the ambient temperature of the infrared sensor 5 is small, a new correction coefficient setting process may not be performed. If the correction coefficient is set when the amount of change in the ambient temperature of the infrared sensor 5 increases as in this example, the processing load on the processing device 6 can be reduced. Note that the ambient temperature of the infrared sensor 5 can be easily determined from the output of this temperature sensor by providing a temperature sensor in the vicinity thereof. In particular, when it is used in a usage form mounted on a vehicle like the vehicle surroundings monitoring device 1, the ambient temperature of the infrared sensor 5 may be determined using the output from a temperature sensor already installed in the vehicle. it can.

The correction processing unit 13 includes a pixel to be corrected among the pixels of the infrared sensor 5, that is, the difference calculation unit 12
The pixel value from the infrared sensor 5 is corrected using the correction coefficient for the pixel for which the correction coefficient is set by, and the corrected value is output as an output value. In addition, the correction processing unit 1
3 outputs the pixel value from the infrared sensor 5 as an output value for the pixel for which the correction coefficient is not set. The output value of each pixel output from these correction processing units 13 is
The image signal is supplied to the display device 3, and the image is displayed on the display device 3.

As a method of correcting the pixel value of the correction target pixel by the correction processing unit 13, a method of correcting the correction target pixel with the latest correction coefficient set by the difference calculation unit 12 is also conceivable. Considering that the pixel value from 5 may suddenly change due to noise mixing or the like, in order to eliminate such a sudden factor, the correction coefficient set in the past for the correction target pixel is excluded. Is stored in the storage device 7, and when the correction target pixel is corrected, the correction coefficient set in the past for the correction target pixel is read from the storage device 7, and the correction coefficient set in the past is read. It is desirable to correct the pixel value of the correction target pixel by an arithmetic process including a simple average, a weighted average, or a moving average.

Here, the number of times the correction coefficient is set for the pixel to be corrected is n, and the correction coefficient set each time is k.
1, k2 ... kn, the simple average is (k1 + k2 +
... + kn) / n. Also, the latest correction coefficient k
When the weight w is set to n, the weighted average is (k1 +
k2 + ... + w · kn) / n. Further, the moving average from the latest correction coefficient kn to α times before (where 1 <α <n) is {k (n−α + 1) + k (n−α + 2).
+ ... + kn} / α.

As described above, the pixel value of the correction target pixel is corrected by the arithmetic processing including the simple average, the weighted average, or the moving average with the correction coefficient set in the past for the correction target pixel. In such a case, even if there is a sudden output change due to noise mixing or the like, it is possible to perform an appropriate correction process that is not affected by such a sudden factor.

In order to eliminate the fluctuation component peculiar to each pixel of the infrared sensor 5, an initial correction coefficient for correcting the fluctuation component peculiar to that pixel is preset for each pixel of the infrared sensor 5. The correction coefficient set in the past for the correction target pixel is stored in the storage device 7, and the correction target pixel is set to the correction target pixel when the correction target is corrected. On the other hand, the preset initial correction coefficient and the correction coefficient set in the past for the pixel to be corrected are read from the storage device 7 and include a simple average, a weighted average, or a moving average with these correction coefficients. It is more desirable to correct the pixel value of the correction target pixel by arithmetic processing.

Here, when the initial correction coefficient preset for the pixel to be corrected is k0, the number of times the correction coefficient is set is n, and the correction coefficients set at each time are k1, k2 ... kn, the simple average is calculated. Is (k0 + k1 + k2 + ... + kn) /
It is calculated by (n + 1). When the weight w is set to the latest correction coefficient kn, the weighted average is (k0 + k1
+ K2 + ... + w · kn) / (n + 1). In addition, α times before (where 1 <α
The moving average up to <n) is {k0 + k (n-α + 1) +
It is obtained by k (n-α + 2) + ... + kn} / (α + 1).

As described above, the simple average, the weighted average, or the moving average of the initial correction coefficient set in advance for the correction target pixel and the correction coefficient set in the past for the correction target pixel is included. When the pixel value of the correction target pixel is corrected by the arithmetic processing, even when there is a sudden output change due to noise mixing or the like,
It is possible to eliminate such a sudden factor and also eliminate the fluctuation component peculiar to the pixel to be corrected, and perform more appropriate correction processing.

Here, how the image is corrected by the correction processing of the processing device 6 as described above will be briefly described with reference to FIGS. 4 to 6.

FIGS. 4 to 6 show images of the rear of the vehicle including the human body when the human body is present behind the vehicle, taken in time series in the order of FIG. 4, FIG. 5, and FIG. Figure 4
4A, FIG. 5A and FIG. 6A respectively show images of images before the correction processing by the processing device 6, and FIG.
(B), FIG. 5 (b), and FIG. 6 (b) respectively show images of the image after the correction processing by the processing device 6.

First, when an image of the rear of the vehicle including a human body as shown in FIG. 4 (a) is taken, of the pixels forming this image, the pixel of the background, that is, the pixel value is predetermined. Pixels having a threshold value Vp or less are set as correction candidates. Figure 4
In the image of the image shown in (a), the hatched area corresponds to the background, and each pixel in this area is a correction candidate. Here, if the pixels A, B, and C in the entire image need to be corrected, then among the pixels A, B, and C that need to be corrected, only the pixel A is within the area corresponding to the background. Since it is a pixel, this pixel A is set as a correction candidate pixel.

Next, among the correction candidate pixels (pixels in the area corresponding to the background), the pixel value of which is different from the reference value Vr, that is, the pixel A in the example shown in FIG. 4A is corrected. It is recognized as a target pixel. Then, for this pixel A, the difference between the pixel value and the reference value Vr is set as a correction coefficient, and the correction process for pixel A is performed using this correction coefficient. At this time, the pixels B and C are pixels that need to be corrected, but since they are pixels in the region where the human body is imaged, correction processing is not performed on these pixels B and C at this stage.

By correcting the pixel value of the pixel A by the above processing, an image as shown in FIG. 4B is displayed on the display device 3. Next, when the position of the human body moves and the image shown in FIG. 5 (a) is captured, an area corresponding to the background shown by hatching in FIG. 5 (a) among the pixels forming this image. Each pixel inside is set as a correction candidate. Here, the pixels B and C that are not corrected by the previous correction processing are the pixels that need to be corrected. Then, among the pixels B and C that need to be corrected, only the pixel B is a pixel in the area corresponding to the background, and thus the pixel B is a correction candidate pixel.

Next, among the correction candidate pixels (pixels in the area corresponding to the background), the pixel value of which is different from the reference value Vr, that is, the pixel B in the example shown in FIG. It is recognized as a target pixel. Then, for this pixel B, the difference between the pixel value and the reference value Vr is set as a correction coefficient, and the correction process for pixel B is performed using this correction coefficient. At this time, the pixel C is a pixel that needs to be corrected, but since the pixel C is a pixel in the region where the human body is imaged, the correction process is not performed on this pixel C at this stage.

The pixel value of the pixel B is newly corrected by the above processing, so that an image as shown in FIG. 5B is displayed on the display device 3.

Next, the position of the human body is further moved, and as shown in FIG.
When the image shown in (a) is picked up, among the pixels forming this image, each pixel in the area corresponding to the background shown by hatching in FIG. 6 (a) is a correction candidate. Here, the pixel C that is not corrected by the previous correction processing is the pixel that needs correction. Then, the pixel C that needs this correction
Is a pixel in the area corresponding to the background, this pixel C
Is a correction candidate pixel.

Next, of the correction candidate pixels (pixels in the area corresponding to the background), the pixel value of which is different from the reference value Vr, that is, the pixel C in the example shown in FIG. 6A is corrected. It is recognized as a target pixel. Then, for this pixel C, the difference between the pixel value and the reference value Vr is set as a correction coefficient, and the correction process for pixel C is performed using this correction coefficient.

The pixel value of the pixel C is newly corrected by the above processing, so that the correction processing is performed on all of the pixels A, B, and C that need to be corrected, and the display device 3 is displayed in FIG.
A clear image as shown in (b) is displayed. After that, by repeatedly performing the above-described processing, it is possible to perform the correction processing on the pixels that need to be corrected while always capturing the necessary image such as the image behind the vehicle, and the display device 3 displays a clear image. Can be displayed.

Here, a flow of a series of processes in the vehicle rear monitoring device 1 to which the present invention as described above is applied will be described with reference to the flowchart of FIG.

When the vehicle rear monitoring device 1 is activated, first, in step S1, the initial correction coefficient of each pixel of the infrared sensor 5 of the infrared camera 2 is read from the storage device 7. The initial correction coefficient is set in advance and stored in the storage device 7 as a correction component for correcting a variation component peculiar to each pixel of the infrared sensor 5.

When the reading of the initial correction coefficient is completed, next, in step S2, an image of the surroundings of the vehicle is captured under the control of the processing unit 6 of the infrared sensor 2.
Specifically, for each pixel of the infrared sensor 5, a pixel value according to the amount of incident infrared radiation is output and supplied to the processing device 6.

Next, in step S3, the processing device 6
The correction candidate specifying unit 11 determines whether the pixel value output from the infrared sensor 5 is less than or equal to a predetermined threshold value Vp. Then, a pixel having a pixel value equal to or smaller than the threshold value Vp is specified as a correction candidate pixel, and among the pixels specified as the correction candidate, for the pixel having a pixel value different from the reference value Vr, the processing device 6 is executed in step S4. Difference calculation unit 1
2, the difference between the pixel value and the reference value Vr is obtained, and this difference is set as a new correction coefficient for the pixel. On the other hand, no new correction coefficient is set for a pixel whose pixel value exceeds the threshold value Vp.

Next, in step S5, it is determined whether or not the pixel value determination as described above has been performed for all pixels. Then, if there is a pixel for which the pixel value has not been determined, the process returns to step S3, and the processes after step S3 are repeated.

On the other hand, when the determination of the pixel values for all the pixels is completed, next, in step S6, the processing device 6
The correction processing unit 13 performs the correction processing using the correction coefficient on the pixel value of the pixel for which the correction coefficient is set,
The corrected value is output as the output value. The correction process using this correction coefficient is performed by, for example, an arithmetic process including a simple average, a weighted average, or a moving average of the initial correction coefficient, the correction coefficient set in the past, and the correction coefficient newly set. Further, at this time, with respect to the pixel for which the correction coefficient is not set, the pixel value is directly output without being corrected.

Next, in step S7, the processing device 6
The display device 3 is driven based on the output value (image signal) of each pixel output from the correction processing unit 13 of
The image captured by the infrared camera 2 is displayed at. Then, in step S8, it is determined whether or not the image pickup by the infrared camera 2 is finished, and if the image pickup by the infrared camera 2 is not finished,
Returning to step S2, the same processing is repeated. On the other hand, when the image capturing by the infrared camera 2 is completed, the series of processes in the vehicle rear monitoring device 1 to which the present invention is applied is completed.

As described above, in the vehicle surroundings monitoring apparatus 1 to which the present invention is applied, among the pixels of the infrared sensor 5, the pixels whose pixel value is less than or equal to the predetermined threshold value Vp are the pixels of which the background in the image is picked up. Assuming that this pixel is a correction candidate, a correction coefficient is set for a pixel that outputs a pixel value different from the reference value Vr among the correction candidate pixels, and the pixel value correction processing is performed. Therefore, it is possible to perform the correction process for the pixel values of the pixels that need to be corrected at the same time while continuing to capture the image. Therefore, the image showing the situation around the vehicle can be constantly displayed on the display device 3 as a high-definition image, and the occupant can appropriately judge the situation around the vehicle.

[Brief description of drawings]

FIG. 1 is a block diagram showing an overall configuration of a vehicle surroundings alerting device to which the present invention is applied.

FIG. 2 is a functional block diagram of a processing device included in the infrared camera of the vehicle surroundings warning device.

FIG. 3 is a diagram illustrating a method of identifying a correction candidate pixel.

FIG. 4 is a diagram showing an example of an image captured by the infrared camera, (a) shows an image of an image before correction processing by the processing device, and (b) shows after correction processing by the processing device. The image of the image is shown.

FIG. 5 is a diagram showing another example of an image captured by the infrared camera, (a) shows an image of an image before correction processing by the processing device, and (b) shows correction processing by the processing device. The image of a later image is shown.

6A and 6B are diagrams showing still another example of an image captured by the infrared camera, FIG. 6A shows an image of an image before correction processing by the processing device, and FIG. 6B shows correction by the processing device. The image of the image after a process is shown.

FIG. 7 is a flowchart showing a flow of a series of processes performed by the vehicle surroundings warning device.

[Explanation of symbols]

1 Vehicle surroundings warning device 2 infrared camera 3 display devices 5 infrared sensor 6 processing equipment 7 storage device 11 Correction candidate specifying unit 12 Difference calculator 13 Correction processing unit

Claims (6)

[Claims] 1. An infrared sensor having a plurality of pixels, each pixel outputting a pixel value according to an incident infrared ray, and each pixel value output from the infrared sensor is compared with a predetermined threshold value to obtain a pixel value. A pixel whose value is less than or equal to the threshold value is specified as a correction candidate, and for the specified correction candidate pixel whose pixel value is different from a predetermined reference value, the difference between the pixel value and the reference value is a correction coefficient. The correction coefficient setting means for setting the correction coefficient and the pixel for which the correction coefficient is not set by the correction coefficient setting means outputs the pixel value output by the infrared sensor as an output value, and the correction coefficient is set by the correction coefficient setting means. Image output means for outputting a value obtained by correcting the pixel value output by the infrared sensor using the correction coefficient, as an output value. External camera. 2. The infrared camera according to claim 1, wherein the correction coefficient setting means performs the correction coefficient setting process for each image output cycle. 3. The infrared camera according to claim 1, wherein the correction coefficient setting means performs the correction coefficient setting process when the amount of change in the output image becomes large. 4. The infrared camera according to claim 1, wherein the correction coefficient setting means performs the correction coefficient setting process when the amount of change in the ambient temperature of the infrared sensor becomes large. 5. The pixel for which the correction coefficient is set by the correction coefficient setting means by the image output means by a calculation process including a simple average, a weighted average, or a moving average with a correction coefficient set in the past. The infrared camera according to claim 1, wherein a pixel value is corrected. 6. The correction coefficient setting means causes the image output means to perform a calculation process including a simple average, a weighted average, or a moving average of a preset initial correction coefficient and a previously set correction coefficient. The infrared camera according to claim 1, wherein a pixel value of a pixel for which a correction coefficient is set is corrected.