JP2001154646A - Infrared picture display device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an infrared picture display device capable of preventing the reduction of the visibility of a display picture even when an object having a temperature different from that of a background moves in the background where a change is small and when the side of the infrared picture display device moves with respect to the object being standing still. SOLUTION: This device is provided with a histogram preparing part preparing the histrogram of an original picture by dividing it into plural temperature ranges and a gradation transformation part which prepares a histogram posterior to a gradation transformation by assigning the number of signal levels posterior to the gradation transformation in accordance with numbers of respective signal levels of the divided plural temperature ranges and which performs a transformation with a signal level posterior to the gradation transformation by referencing the signal level of the original picture stored in a frame memory as a key.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an infrared image display device, and more particularly to an infrared image display device when an object having a temperature different from the background temperature moves in a background having a small temperature change or an object having a temperature different from the background temperature. The present invention relates to an infrared image display device capable of preventing a decrease in visibility of a display image when the infrared image display device moves with respect to the infrared image display device.

[0002] All objects emit infrared radiation of an intensity which is uniquely related to the temperature of the object itself. Then, the wavelength of the emitted infrared light depends on the substance forming the object.

An infrared image display device receives infrared light emitted from an object as one-dimensional information by an infrared detector, performs photoelectric conversion, converts an analog electric signal, which is a photoelectric conversion output, into a digital signal, The variation of the output voltage at the same temperature as the detection sensitivity between the plurality of infrared detection elements constituting the infrared detector, that is, the offset, is corrected, and the display accuracy is converted (this is referred to as “gradation conversion”.
Hereinafter, it is referred to as "gradation conversion" consistently. ), Perform a predetermined sweep by the monitor, convert it to a two-dimensional image,
Display an image with visible light.

As described above, an infrared image display device detects and displays infrared rays radiated from an object, and thus is significantly different from an image display device that captures and displays visible light using a camera or the like. Regardless of the object, the object can be recognized.

[0005] In addition, radar and sonar are very different from each other, and do not emit electromagnetic waves, sound waves, or ultrasonic waves in order to recognize the object, but passively capture the infrared rays emitted by the object itself. Has the great advantage that the object can be recognized without noticing the object side of the recognition target.

Further, detecting an infrared ray emitted from an object to recognize the object means that the object can be recognized in a non-contact manner. Therefore, the object cannot be approached or approached. Even when the thing itself is dangerous, the object can be easily recognized.

[0007] The infrared detecting element used in the infrared image display apparatus has a 3-5 µm band from the detection wavelength range.
It is roughly classified into a band of about 12 μm. Platinum silicon (PtSi), indium tin (I)
nSn), mercury cadmium tellurium (HgCdTe) or the like is used, and mercury cadmium tellurium is mainly used for the sensing element in the 8 to 12 μm band.

[0008] In particular, mercury-cadmium-tellurium, which is an alloy of mercury, cadmium, and tellurium, has attracted attention because a high-sensitivity sensing element can be realized in a wide wavelength range by changing the composition ratio thereof.

Therefore, the infrared image display device can acquire temperature information over a wide temperature range from extremely low temperature to high temperature.

Taking advantage of such many features, infrared image display devices are used in a very wide range of fields including public and private sectors such as surveillance cameras, thermography, remote sensing, and forward monitoring devices for moving objects such as aircraft and vehicles. Used in Of course, it is also used in the field of national defense, and recently, its use in the fields of medicine and sports physiology has attracted attention.

The image displayed by the infrared image display device includes an image having a small change in temperature information and an object having a temperature different from the background temperature moving in a background having a small temperature change.
There is an image in which the temperature information may change because the infrared image display device moves to an object having a different temperature from the background. In the former case, there is no change in the visibility of the infrared image because the change is originally small, but in the latter case, there is a problem that the visibility is reduced.

Therefore, when an object having a temperature different from the background temperature moves in a background having a small change, or when the infrared image display device side moves to an object having an object temperature different from the background temperature, There is a need for an infrared image display device capable of preventing a reduction in the visibility of a displayed image.

[0013]

2. Description of the Related Art FIG. 14 shows a general configuration of an infrared image display device.

In FIG. 14, reference numeral 1 denotes an optical system including a lens system for condensing infrared rays to be received and a mechanism for scanning the condensed infrared rays. An infrared detector 3 for outputting an analog electric signal, and an amplifier for amplifying the output of the infrared detector 2 to a predetermined level in order to secure a signal-to-noise ratio of the analog electric signal output from the infrared detector 2 4 is the amplifier 3
Analog-to-digital converter (A / D in FIG. 1) that converts the analog output of D into a digital signal.
“D” is an initial of “Degital”, and “D” is an initial of “alog”. ), 5 is a frame memory for storing the output of the analog-to-digital converter 4 during image processing (in the figure,
Although labeled as "frame memory", they are the same. ) And 6 are histogram creating units for counting the number of appearances of the output level of the analog-to-digital converter 4 (analog signals converted into digital codes) and creating a histogram of the original image. A histogram conversion unit 9 for creating a histogram in which the histogram of the original image created by the histogram creation unit 6 is converted in accordance with the gradation to be displayed, and converting the original image into an image in which the tone is converted; 9
Is a digital-to-analog converter (abbreviated as “D / A” in the figure) for inversely converting the grayscale-converted digital signal output from the histogram converter 7 into an analog signal. "D" and "A"
Is abbreviated as in the case of the analog / digital converter. Reference numeral 10 denotes a monitor that performs a predetermined sweep on the analog signal output from the digital / analog converter 9 and displays the same as a two-dimensional image.

As described above, since the dynamic range of infrared rays received by the optical system 1 is extremely wide, the accuracy of the analog / digital converter 4 needs to be about 12 bits.

On the other hand, it is the human eye that recognizes temperature information and the presence or absence of an object based on the image displayed on the monitor 10.
Since the resolution of the human eye is about 50 dB as a signal-to-noise ratio, it is about 8 bits.

Therefore, the monitor 10 only needs to display the temperature information and the accuracy required for recognizing the object.
That is, gradation conversion is performed using the memory 5, the histogram creation unit 6, and the histogram conversion unit 7.

FIG. 15 is a diagram for explaining the conventional image processing, taking an example of an image composed of an object in a high temperature range and a background in a normal temperature range.

FIG. 15A shows the creation of a histogram of an original image. The vertical axis represents the number of pixels, and the horizontal axis represents the signal level represented by 12 bits. Since the number of signal levels that can be represented by 12 bits is 4096, the horizontal axis indicates 0 to 4095.
Is touched.

Here, since an image composed of an object in a high temperature range and a background in a normal temperature range is shown as an example, the histogram is 2
Divided into two mountains. Since the background in the normal temperature range occupies most of the field of view, the mountain in the normal temperature range is usually larger. This histogram is created by the histogram creating unit 6 in FIG.
Is forwarded to

The histogram converter 7 cuts the signal level where the number of pixels in the histogram of the original image is less than a predetermined number (expressed as "threshold" in the figure), and the histogram is not cut. Only the remaining signal levels are rearranged, and the total number N of signal levels remaining without being cut is obtained. This is shown in FIG.

Next, as shown in FIG. 15C, the signal level N on the horizontal axis is evenly allocated to the 8-bit display accuracy. Since the signal level that can be represented by 8 bits is 256, 0 to 255 are added to the vertical axis. Thus, the gradation of the 12-bit original image is compressed to 8 bits, and stored in a histogram conversion memory (not shown in FIG. 14) in the histogram conversion unit with the signal level of the pixel of the original image as an address. You.

Thereafter, the image data of the same frame as the histogram stored in the histogram conversion memory, that is, the digitized signal level is read from the frame memory 5 of FIG. 14, and the read signal level is used as an address. By reading the 8-bit gradation image data stored in the histogram conversion memory,
The original image is converted into an 8-bit image.

That is, the histogram conversion memory is a conversion table for reading out 12-bit gradation and 8-bit gradation. In short, the signal level of the original image is given, and the signal level of the corresponding display image is obtained.
This is represented graphically and shown in FIG.

Therefore, the histogram of the image converted into 8 bits becomes a histogram distributed among 256 signal levels, as shown in FIG. This is shown in FIG.
In the histogram of 5 (b), the maximum signal level number N
Is converted to 255, and the shape of the histogram is preserved before and after the gradation conversion.

Then, as described above, the 8-bit gradation image data in which the signal level of the 12-bit original image is stored as an address in the histogram conversion memory is converted into 12-bit gradation image data.
Since the signal level of the original bit image is read as an address and gradation conversion is performed, the image whose gradation is converted to 8 bits becomes an image faithful to the original image.

In the above description, the case where there is a high-temperature object in the background is described. However, in the above-described image processing, even when there is a low-temperature object in the background, there are a high-temperature object and a low-temperature object in the background. Needless to say, it can be applied to the case.

A typical example of a high-temperature object is an automobile running an engine, a typical example of a low-temperature object is a clear sky, and an example of a moving low-temperature object is a clear sky. There is a mobile windshield that reflects the infrared light.

[0029]

However, in order to improve the visibility of an image whose gradation has been converted to 8 bits, it is necessary that the original image itself has little change.

FIG. 16 is a diagram for explaining a problem in the conventional image processing.

FIG. 16A shows an original image, a histogram of the original image, and a histogram after gradation conversion when there is no high-temperature object in the background at room temperature. For example, you can imagine a landscape such as a mountain or a field.

On the other hand, FIG. 16B shows an original image, a histogram of the original image, and a histogram after gradation conversion when there is a high-temperature object in the background at room temperature. For example, it may be assumed that the car is stationary or running with the engine running against a background of mountains or fields.

When a histogram is created for a 12-bit signal level when no hot object exists in the original image,
The histogram is as shown in the histogram of the original image in FIG.
When this is converted into an 8-bit histogram by the method described above, all 256 signal levels of the histogram are occupied only by the signal level indicating the background in the normal temperature range.

On the other hand, if a histogram is created for a 12-bit signal level when a high-temperature object is present in the original image, the histogram is as shown in the original image in FIG. Here, since the temperature distribution of the background portion is not related to the presence or absence of a high-temperature object, the histogram of the background portion in the histogram of the original image remains unchanged.

When the histogram of the original image in the presence of the high-temperature object is converted into an 8-bit histogram by the method described above, 256 signal levels are shared between the background in the normal temperature range and the high-temperature object. .

Therefore, in the histogram after gradation conversion, the number of levels occupied by the background in the normal temperature range changes depending on the presence or absence of a high-temperature object.

Considering the case where the high-temperature object is stopped or moves slowly in the background, no change occurs in the histogram after gradation conversion for the original image in FIG. Because the histogram represents the temperature distribution in the original image, not the original image itself,
This is because a change in the position of the hot object does not affect the histogram.

Therefore, when the high-temperature object is stopped or moves slowly in the background, the histogram after gradation conversion in FIG. 16B does not change, and the visibility of the displayed image is good. Kept in state.

Next, consider a case where a high-temperature object moves at high speed in the background. A particularly prominent case is a case where a high-temperature object moves at high speed in the background at a close distance of the infrared image display device.

In this case, the original image changes in a short time, such as a state where the high-temperature object is not in the original image, a state where the high-temperature object has entered the original image, and a state where the high-temperature object has left the original image. .

First, a histogram after gradation conversion when a high-temperature object is not present in the original image is shown in the rightmost diagram in FIG.

When a high-temperature object enters the original image at a high speed, the histogram after gradation conversion changes instantaneously, and changes as shown in the rightmost diagram in FIG. That is,
The number of levels in the background histogram is compressed, and the background histogram is shifted to the left.

At the next moment, the high-temperature object comes out of the original image at a high speed, and the histogram after gradation conversion changes within a moment, as shown in the rightmost diagram in FIG. Become. That is, the number of signal levels assigned to the histogram of the background portion increases, and the histogram of the background portion is shifted to the right.

The shift of the histogram of the background portion to the left and right is clearly shown by the auxiliary line drawn on the peak portion of the histogram in the rightmost diagrams of FIGS. 16 (a) and 16 (b).

That is, when the high-temperature object moves at high speed in the original image, the number of signal levels allocated to the background in the normal temperature range changes abruptly in the histogram after gradation conversion.
The appearance of the background portion changes rapidly with the movement of the high-temperature object, and the displayed image is blurred.

In this state, the change in the histogram of the background portion is masked, and the visibility of the movement of the hot object is reduced.

Such a situation can occur even when a low-temperature object moves rapidly in the background.

The fact that monitoring is performed by the infrared image display device is, of course, not for the purpose of confirming only the background, but also for confirming the presence / absence of an abnormal object that enters and exits. Also, the fact that the infrared image display side monitors the front while moving at high speed means that
It makes sense to find abnormal objects in the constantly changing original image.

Therefore, when the abnormal object moves at high speed in the original image, the visibility of the infrared image display device is not allowed to decrease.

In view of the above, the present invention provides an infrared image display apparatus for a case where an object having a temperature different from the background moves at a high speed in a background with little change, or for an object having a temperature different from the background. It is an object of the present invention to provide an infrared image display device capable of preventing a reduction in visibility of a display image even when the side moves at high speed.

[0051]

The first means of the present invention is as follows.
A histogram of signal levels appearing in the original image is created by dividing it into a plurality of temperature regions, and the number of signal levels after gradation conversion is assigned according to the number of signal levels in each of the plurality of divided temperature regions. This is a technique for converting between a signal level and a signal level after gradation conversion.

According to the first means of the present invention, a histogram of an original image is created by dividing it into a plurality of temperature ranges, and after the gradation conversion according to the number of signal levels in each of the plurality of divided temperature ranges. Since the number of signal levels is assigned, there is no change in the number of signal levels assigned to the background after gradation conversion, for example, whether or not there is a high-temperature object in the original image.
Similarly, whether or not there is a low-temperature object in the original image does not change the number of signal levels assigned to the background portion after gradation conversion.

Therefore, even when the abnormal object moves at high speed in the original image, the visibility of the displayed image does not decrease.

The second means of the present invention is to average the signal level histograms appearing in the original image over a plurality of frames, assign the signal levels after gradation conversion, and compare the signal levels of the original image with the signal after gradation conversion. This is a technique for converting between levels.

According to the second aspect of the present invention, the histogram of the original image is averaged over a plurality of frames, and the signal level after gradation conversion is assigned, and the signal level of the original image and the signal level after gradation conversion are determined. Therefore, even if an abnormal object moves at high speed in the original image, the histogram of the background portion in the histogram after gradation conversion is not rapidly affected by the presence or absence of the abnormal object.

Therefore, even when the abnormal object moves at high speed in the subject, the visibility of the displayed image does not decrease.

The third means of the present invention calculates the difference between the histogram calculated and stored one frame before and the histogram of the current original image, weights the difference differently depending on the temperature range, The weighted difference is added to the histogram stored one frame before, and the histogram obtained by adding the weighted difference and the histogram stored one frame before is assigned the number of signal levels after gradation conversion. Is a technique for converting between the signal level of an original image and the signal level after gradation conversion.

According to the third means of the present invention, a difference between the histogram stored one frame before and the histogram of the current original image is obtained, and a large weighting factor is applied to the normal temperature portion with respect to the difference. The high temperature portion is multiplied by a small weighting factor, the weighted difference is added to the histogram stored one frame before, and the histogram obtained by adding the weighted difference and the histogram stored one frame before is added to the histogram. On the other hand, since the number of signal levels after gradation conversion is assigned, even if an abnormal object moves at high speed in the original image, the histogram of the background part in the histogram after gradation conversion is rapidly affected by the presence or absence of the abnormal object. Nothing.

[0059]

FIG. 1 shows a hardware configuration of a first embodiment of the present invention.

In FIG. 1, reference numeral 1 denotes an optical system including a lens system for condensing infrared rays to be received and a mechanism for scanning the condensed infrared rays. An infrared detector 3 for outputting an analog electric signal, and an amplifier for amplifying the output of the infrared detector 2 to a predetermined level in order to secure a signal-to-noise ratio of the analog electric signal output from the infrared detector 2 4 is the amplifier 3
Is an analog-to-digital converter that converts the analog output of the device into a digital signal.

Reference numeral 51 denotes a first selector (in the figure, referred to as "first SEL") which sequentially outputs the output of the analog-to-digital converter 4 to three output terminals.
In the figure, they are similarly indicated. “SEL” is “SELECTER”
Is an abbreviation with the first three letters. ), 52 are first frame memories for storing the signal levels of the pixels for one frame (in the figure, since the random access memory constitutes the frame memory, it is denoted as "RAM"). In the following, the memory using the random access memory will be indicated in the same manner.
Abbreviation for "ndom Access Memory", which is very commonly used. ) And 53 are second frames for storing the signal levels of the pixels for one frame.
A memory 54 is a third frame memory for storing a signal level of a pixel for one frame, and 55 is a third frame memory from the first frame memory 52, the second frame memory 53 and the third frame memory 54. A second selector for selecting any of the image data to be read, the first selector 51, the first frame memory 52, the second frame memory 53, the third frame memory 54, and the second The selector 55 causes the frame memory 5 of FIG.
Is configured.

The first to third frame memories store the signal level of the original image, that is, the image data, using the pixel address of the original image as an address. The signal level of the original image is read using the pixel address of the original image as an address. That is,
In the figure, address lines to the first to third frame memories are omitted.

Reference numeral 61 denotes a histogram creation memory for creating a histogram of signal levels appearing in the original image, 62 denotes an adder for adding 1 to the number of appearances of the signal level read from the histogram creation memory 61,
Reference numeral 63 denotes a buffer gate. The histogram creation memory 6, the adder 62 and the buffer gate 63 constitute the histogram creation unit 6 in FIG. 14.

Reference numeral 71 denotes a digital signal processor (in the figure, "DSP" for executing the histogram conversion. In the following description, the same applies in the figure. "DSP" is the head of "Digital Signal Processor". An abbreviation using characters is becoming a general abbreviation.) 72 is a histogram storage memory for receiving and storing the histogram of the original image created by the histogram creation memory 61, and 73 is a digital storage abbreviation. A read-only memory storing a program to be executed by the signal processor 71 (in the figure,
Abbreviated as "ROM". Hereinafter, the same is applied in the figures. Note that “ROM” is an abbreviation that takes the initials of “Read Only Memory” and is a very common abbreviation.
When executing a program, instead of reading out one step at a time from the read-only memory, a certain range of program steps is read out from the read-only memory and temporarily stored in the random access memory, and one step is read therefrom. Normally, data is read out and executed one by one, but a random access memory for that purpose is not shown. ) And 74 are digital signal processors 71
Writes a histogram obtained by performing gradation conversion from the histogram stored in the histogram storage memory 72 using the signal level of the original image as an address, and reads the signal level of the original image output from the second selector 55 as an address. The dual port random access memory as a table (in the figure, it is described as "DP-RAM". In the following description, it is similarly shown in the figure.
“P-RAM” is an abbreviation for “Dual Port Random Access Memory”, which is commonly used. 14), the digital signal processor 71, the histogram storage memory 72, the read-only memory 73, and the dual port random access memory 74 constitute the histogram conversion unit 7 in FIG.

However, as will be described in detail later, the program stored in the read-only memory 73 is different from the conventional technique.

Reference numeral 9 denotes a digital-to-analog converter for converting data read from the dual-port random access memory 74 into an analog signal.

Reference numeral 10 denotes a monitor for displaying the output of the digital / analog converter 9 as a two-dimensional image by a predetermined sweeping method.

Here, since the dynamic range of infrared rays received by the optical system 1 is extremely wide, the analog-to-digital converter 4 needs approximately 12 bits for accurate digital conversion.

On the other hand, it is the human eye that recognizes the temperature information and the object by the image displayed on the monitor 10, and the resolution of the human eye is about 8 bits.

Therefore, the monitor 10 only needs to display the temperature information and the accuracy required for recognizing the object.
The memory, the histogram creation unit, and the histogram conversion unit perform 8-bit gradation conversion as described in detail below.

Next, since the dual port random access memory 74 functions as a gradation conversion table from 12 bits to 8 bits, the capacity of the dual port random access memory 74 is 12 bits × 8.
Is a bit. The address of the digital signal processor 71 is 12 bits.

Further, in a normal infrared image display device, the number of display pixels is 320 in the horizontal direction and 240 in the vertical direction.
The capacity is 17 bits, and the data line is 17 bits. Therefore, the adder 62 can handle 17 bits.

The capacity of the first to third frame memories and the capacity of the histogram storage memory 72 are 17 bits × 12 bits.

The creation of the histogram is performed as follows.

First, since there must be no offset in the creation of the histogram, the operation is started after the histogram creation memory 61 is reset once after the power is turned on.

The analog-to-digital converter 4 supplies a specific signal level of the 12-bit signal level to the histogram creation memory 61 as an address at a specific time. Since the histogram creation memory 61 has been powered on and reset as described above, 0 is stored in the address at first. This 0 is read out and added to 1 in the adder 62, and the added 1 is stored in the corresponding address of the histogram creation memory 61.

Next, if the same signal level as that of the previous output is output from the analog-to-digital converter 4 and supplied to the histogram creation memory 61 as an address, the result of adding 1 again is stored in the address. Is done.

On the other hand, if a different signal level is supplied as an address from the analog / digital converter 4, 1 is added to a different address in the histogram creation memory 61 and stored.

In this way, the number of appearing pixels is stored in the histogram creation memory 61 for each output level of the analog-to-digital converter 4, and a histogram of the original image is created.

Since the histogram of the original image is sequentially created for each pixel in the frame, the creation of the histogram of the original image and the writing of the signal level of the original image to the frame memory are performed simultaneously.

In the histogram creation memory 61, 1
When a histogram is created for the images in the frame memory, the created histogram of the original image is transferred to and stored in the histogram storage memory 72.

The digital signal processor 71 operates in accordance with a program stored in the read-only memory 73,
An operation is performed on the histogram of the original image written in the histogram storage memory 72, and the operation result is written into the dual port random access memory 74 using the signal level of the original image as an address. The data written at this time is gradation data compressed to 8 bits.
Note that this operation is a characteristic operation of the digital signal processor 71 in the first embodiment of the present invention, and will be described later in detail.

[0083] Dual port random access
The 8-bit gradation data written in the memory 74 is 3 bits.
Read out from the dual port random access memory 74 using the signal level of the original image read out from one of the first to third frame memories as an address, and It is supplied to the converter 9. Therefore, the data supplied to the digital-to-analog converter 9 is a signal level converted into a gradation corresponding to the pixel address, that is, image data.

This is converted into an analog signal in the digital / analog converter 9, supplied to the monitor, subjected to a predetermined sweep, and displayed as a two-dimensional image.

FIG. 2 is a diagram showing the above procedure as a diagram showing a processing procedure according to the first embodiment of the present invention.

FIG. 2A shows a frame.
A frame is composed of two fields. here,
The first to fourth frames are illustrated.

FIG. 2B shows valid data of the original image data.

The valid data of the image does not fill the field, but the time required for processing before and after the switching of the field is taken, and the valid data is given to the remaining time.

FIG. 2C shows the storage of valid data in the frame memory. The acquired valid data is sequentially stored in the frame memory for each frame. Here, the valid data of the pixels of the first frame is illustrated as being stored in the first frame memory. Therefore, the subsequent valid data is sequentially stored in one of the three frame memories.

At the same time that the signal levels of the pixels of the first frame are stored in the first frame memory, the signal levels representing the effective data of the pixels of the first frame are supplied to the histogram creation memory. , A histogram is successively created. Therefore, a histogram of the signal levels of the first frame is created during the first frame. This is shown in FIG.

The histogram of the image of the first frame created in the memory for creating the histogram is transferred to the memory for storing the histogram during the idle time of the first frame, and is converted into the histogram by the digital signal processor during the time of the second frame. The conversion result is stored in the dual port random access memory in the idle time of two frames. This is shown in FIG.

[0092] Dual port random access
The conversion result stored in the memory is read out using the signal level stored in the first frame memory as an address at the time of the third frame. Thus, the signal level subjected to the gradation conversion is supplied to the digital / analog converter. This is shown in “output data 3” in FIG.

That is, it is the time of the third frame that the signal level of the first frame is subjected to gradation conversion and output. During this time, since the signal level of the original image is continuously input, one frame of memory is required for the frame memory.
If the signal level of the first frame is written in the first frame memory and the signal level of the second frame is written in the second frame memory,
It is necessary to write the signal level of the frame to the third frame memory in order, such as writing the signal level to the third frame memory. This makes it possible to perform gradation conversion by associating the gradation-converted histogram of the first frame with the signal level of the pixels of the first frame. Then, the above operation may be repeated in the fourth and subsequent frames.

The description so far is a rough processing procedure of the first embodiment of the present invention shown in FIGS. 1 and 2,
The features of the first embodiment of the present invention have not been described yet. The features of the processing procedure according to the first embodiment of the present invention will be described with reference to FIG.

FIG. 3 is a diagram for explaining the principle of the first embodiment of the present invention, and shows an example in which an original image includes an object in a low temperature range, a background in a normal temperature range, and an object in a high temperature range. It is.

FIG. 3A shows the creation of a histogram of an original image. The vertical axis represents the number of pixels, and the horizontal axis represents the signal level represented by 12 bits. Since the number of signal levels that can be represented by 12 bits is 4096, 0 to 4095 are added to the horizontal axis.

Here, an image composed of an object in a high temperature range, a background in a normal temperature range, and an object in a low temperature range is shown as an example.
Three peaks appear in the histogram of the original image. Note that a is a signal level defined as a boundary between a low temperature region and a normal temperature region in the original image, and b is a signal level defined as a boundary between a normal temperature region and a high temperature region in the original image.

This histogram is created in the histogram creating section of FIG. 1 and transferred to the histogram converting section.

In the histogram conversion unit shown in FIG. 1, a predetermined number of pixels (FIG.
In (a), it is expressed as “threshold”. ) Is cut to rearrange the histograms, and the number of signal levels remaining without being cut is obtained for each temperature range.

Here, as shown in FIG. 3A, the number of signal levels remaining without being cut is N _{1 in} a low temperature range, N _{2 in} a normal temperature range, and N _{3 in} a high temperature range.

When rearranging the signal levels remaining without being cut, the rearrangement is performed for each of the low temperature range, the normal temperature range, and the high temperature range. In other words, handled histogram of signal levels of N ₁ of the low temperature range, a histogram relating to the signal level of the N ₂ of the normal temperature range, and, independently histogram of signal levels of the N ₃ high-temperature range. This is shown in FIG.

Next, the signal level obtained by performing gradation conversion on the signal level of the histogram shown in FIG.
It is shown as converting to 8-bit, 256-level signal levels. ). At this time, 256 signal levels are allocated to each temperature range according to the number of signal levels in each temperature range.

Here, among the signal levels after the gradation conversion, the signal level at the boundary between the low temperature region and the normal temperature region is set to c, and the signal level at the boundary between the normal temperature region and the high temperature region is set to d.

First, when the variable on the horizontal axis of the two-dimensional coordinates in FIG. 3C is i (i can be an integer of sex), a straight line varying from 0 to 1 with a slope 1 / N _1. 1 (hereinafter referred to as L _1.) to define. That is, the equation of the straight line 1 is: L ₁ = i · (1 / N ₁ ) (1)

Based on the above assumption, the signal level of c is assigned to the low-temperature area, and the display level assigned to the low-temperature area is: display level = c · L ₁ (2)

[0106] Next, define a straight line 2 varies from 0 at a tilt 1 / N ₂ to 1 (hereinafter referred to as L _2.). That is,
The equation of the straight line 2 is: L ₂ = i · (1 / N ₂ ) (3)

According to the above assumption, the signal levels from c to d are assigned to the normal temperature range. Therefore, the display level assigned to the normal temperature range is as follows: display level = (d−c) · L ₂ + c (4)

Similarly, if a straight line 3 which changes from 0 to 1 with a slope of 1 / N ₃ (this is defined as L ₃ ) is defined, L ₃ = i · (1 / N ₃ ) (5) The display level assigned to the area is as follows: display level = (255−d) · L ₃ + d (6)

FIG. 3C shows the distribution of the signal levels after the gradation conversion to the respective temperature ranges. With this, 12
The gradation of the bit original image is obtained by being compressed to 8 bits for each temperature range. This is stored in the histogram conversion memory in the histogram conversion unit using the signal level of the pixel of the original image as an address.

Thereafter, the signal level of the same frame as the image stored in the histogram conversion memory is read out from one of the frame memories of FIG. 1, and the read signal level is stored in the histogram conversion memory as an address. The original image is converted into an 8-bit image by reading the 8-bit gradation data.

That is, the histogram conversion memory is a conversion table for reading 12-bit gradation and 8-bit gradation for each temperature range. FIG. 3D shows a conversion table of the 12-bit original image and the 8-bit processed image expressed in two-dimensional coordinates.

Therefore, the hysteresis of the image converted to 8 bits
The togram becomes a histogram shown in FIG. this
Is the signal in the low temperature range in the histogram of FIG.
Level N₁To c, the signal level N in the normal temperature range_TwoTo (d─
c) shows the signal level N in the high temperature range. _ThreeIs divided into (255-d).
The shape of the histogram before and after gradation conversion
Is saved.

Then, as described above, in the histogram conversion memory, the 8-bit signal level in which the data of the 12-bit original image is stored as the address is converted to the 12-bit signal level.
Since the signal level of the original bit image is read as an address and gradation conversion is performed, the image whose gradation is converted to 8 bits becomes an image faithful to the original image.

Further, as described below, according to the first embodiment of the present invention, even if an abnormal object moves at high speed in the original image, the background portion in the histogram after the gradation change is obtained. Does not fluctuate.

FIG. 4 is a diagram for explaining image processing in the first embodiment of the present invention. For simplicity, FIG. 4 shows only a background in a normal temperature range and an object in a high temperature range. explain.

FIG. 4A shows an original image in the case where a high-temperature object does not exist in the original image, a histogram of the original image, and a histogram after gradation conversion. FIG. 4B shows a case in which a high-temperature object exists. 5 shows an original image, a histogram of the original image, and a histogram after gradation conversion.

Since the original image in the case where there is no high-temperature object has only the temperature information in the normal temperature range, the histogram of the original image in this case is obtained by using the same reference numerals as those in FIG. As shown in the figure, it is created below the signal level b.

This is because, in the histogram after gradation conversion, the normal temperature range is converted to a signal level d or lower as shown in the rightmost diagram in FIG.

On the other hand, even if there is a high-temperature object, the temperature information in the normal temperature range is limited to the signal level b or less and a histogram is created. Therefore, as shown in the middle diagram of FIG. Is independent of the presence of hot objects.

When the high-temperature object is present in the original image, as shown in the middle diagram of FIG. 4B, the temperature information of the high-temperature region is limited to the signal level b or higher, and a histogram is created.

When the histogram of the original image created in this way is subjected to gradation conversion, as described above, gradation conversion is performed independently in the normal temperature range and the high temperature range. It is distributed in the same signal level range without being affected.

Therefore, even if the high-temperature object moves through the background in the normal temperature range at a high speed, the signal level at which the histogram of the background is formed does not change, so that the visibility of the displayed image does not decrease.

Of course, even when a high-temperature object stays for a relatively long time in a room temperature background, the visibility of the displayed image is not affected at all.

Also, no problem occurs in the visibility of the displayed image regardless of whether the original image has the temperature information of the normal temperature range and the low temperature range or the temperature information of the normal temperature range, the low temperature range and the high temperature range.

Up to this point, the first embodiment of the present invention has been sufficiently described, but a description will be given with reference to a flowchart in order to confirm the operation again.

FIG. 5 is a flowchart showing the operation of the first embodiment of the present invention. Hereinafter, the operation of the first embodiment of the present invention will be described along the reference numerals in FIG. Here, the case where the original image has temperature information of the normal temperature range and the high temperature range is taken as an example. The boundary values of the signal levels, the number of signal levels and the equation of the straight line used here are the same as those in FIG.

S1. The histogram creation memory 6 of FIG.
Reset 1 to create a state where 0 is stored in all addresses.

This is an indispensable process at the first stage of the histogram creation.

S2. One is added to the address of the histogram creation memory corresponding to the output of the analog-to-digital converter, and the result is stored in the corresponding address of the histogram creation memory.

As a result, a histogram is sequentially created for each signal level output from the analog-to-digital converter.

S3. It is determined whether the process of step S2 has been completed for all pixels.

Since the number of pixels in one frame is known, it is only necessary to judge whether or not the count value of the stepping counter reaches the above-mentioned number of pixels each time the addition operation is performed.

If it is determined that the addition operation has not been completed for all the pixels (No), the process jumps to step S2, and the processing of steps S2 to S3 is performed until the addition operation is completed for all the pixels. Repeat.

The above is the stage of creating the histogram of the original image, the histogram creation memory 61 and the adder 62 shown in FIG.
Done by

S4. If it is determined in step S3 that the addition operation has been completed for all the pixels (Yes), the histogram of the original image created by the histogram creation memory is transferred to the histogram storage memory 72 of FIG.

S5. For signal levels 0 to b of the original image, signal levels equal to or less than a predetermined number of pixels are cut.

This is because, when the number of gradations is compression-converted, signal levels with a low appearance frequency and low display necessity are deleted, and an image is displayed efficiently within the number of compressed gradations. This is the process to be performed.

S6. The signal levels remaining uncut at the signal levels 0 to b are rearranged, and the total number N ₂ of levels in this signal level range is obtained.

S7. The range of the inclined 1 / N ₂ 0 to 1 to define the straight line 2 take values.

If i is an integer up to 0 and N ₂ and the equation of the straight line is L ₂ , then L ₂ = i (1 / N ₂ ).

S8. When the signal level after gradation conversion corresponding to the signal level b of the original image is d, step S7
Is multiplied by d to convert the display level into 8 bits.

That is, the display level after being converted to 8 bits is: display level = d · L ₂ .

S9. The display level obtained in step S8 is stored in the histogram conversion memory at an address corresponding to the signal level of the original image.

Steps S5 to S9 are the steps of background histogram conversion, and the conversion operation is performed by the digital signal processor 71 of FIG. 1 using the histogram storage memory 72 and the read-only memory 73.

S10. Next, signal levels b to 4095
, The signal level below a predetermined number of pixels is cut.

S11. And rearranging the signal level that remains without being cut by the signal level B～4095, determine the level total number N ₃ of this signal level range.

S12. The range of the inclined 1 / N ₃ 0 to 1 to define the straight line 3 which takes the value.

If i is an integer up to 0 and N ₃ and the equation of the straight line is L ₃ , then L ₃ = i · (1 / N ₃ ).

S13. When the signal level after gradation conversion corresponding to the signal level b of the original image is d, step S12
The display level (255-d) is assigned by the straight line obtained in step (1), and the display level is converted into 8-bit.

That is, the display level after the conversion into 8 bits is as follows: display level = (255−d) · L ₃ + d

S14. The display level obtained in step S13 is stored in the address corresponding to the signal level of the original image in the dual port random access memory.

Steps S10 to S14 are steps of histogram conversion of a high-temperature object. The conversion operation is performed by the digital signal processor 71 using the histogram storage memory 72 and the read-only memory 73.

S15.3 Among the frame memories of the surface, the signal level of the frame memory storing the signal level of the same frame as the frame subjected to the histogram conversion is stored as an address in the histogram conversion memory. The gradation data is read and gradation conversion is performed.

S16. The data read in step S15 is supplied to a digital-to-analog converter to be converted into an analog signal, and output to a monitor for image display.

Steps S15 and S16 are the stages of gradation conversion and display. The gradation conversion is performed by the digital signal processor using the frame memory and the histogram conversion memory.

In this flowchart, the steps of obtaining and converting a histogram in a normal temperature range and the steps of obtaining and converting a histogram in a high temperature range are described.
If there is temperature information of the low temperature range, a step of obtaining and converting the histogram of the low temperature range may be added.

Next, the description moves to the description of the second embodiment of the present invention.

FIG. 6 shows a hardware configuration according to the second embodiment of the present invention.

In FIG. 6, reference numeral 1 denotes an optical system including a lens system for condensing infrared rays to be received and a mechanism for scanning the condensed infrared rays, and 2 denotes an optical system which receives the infrared rays scanned by the optical system 1 and performs photoelectric conversion. , Infrared detector to convert to analog electric signal,
Reference numeral 3 denotes an amplifier for amplifying the output of the infrared detector 2 to a predetermined level in order to secure a signal-to-noise ratio of an analog electric signal output from the infrared detector 2. Analog-to-digital converter.

Reference numeral 51a denotes a third selector for supplying the signal level output from the analog-to-digital converter 4 to any one of the frame memories, and reference numeral 52 denotes a first frame for storing the signal level of the original image for one frame. A memory 53, a second frame memory for storing the signal level of the original image for one frame; 55a, a first frame memory 52;
Alternatively, a fourth selector for selecting the signal level of the original image read from the second frame memory 53, the third selector 51a, the first frame memory 52, the second frame memory 53, and the fourth The frame memory 5 of FIG. 14 is constituted by the selector 55a.

The signal levels of the original image are stored in the first frame memory 52 and the second frame memory 53 using the pixel address of the original image as an address. From the frame memory 53, the signal level of the original image is read using the pixel address of the original image as an address.

Reference numeral 61 denotes a histogram creation memory for sequentially storing the appearance frequency of the signal level of the original image to create a histogram of the original image, and 62 denotes a histogram creation memory 61 for each signal level of the original image. An adder 63 for adding the number and the number of 1 is a buffer gate. The histogram creation memory 6, the adder 62 and the buffer gate 63 of the histogram creation unit 6 shown in FIG.
Is configured.

Reference numeral 71 denotes a digital signal processor for performing histogram conversion; 72a, a histogram storage memory for receiving and transferring the histogram of the original image created by the histogram creation memory; 73a: a digital signal processor 71; The read-only memory 74 storing the program to be executed writes the histogram obtained by the digital signal processor 71 by performing gradation conversion from the histogram stored in the histogram storage memory 72a to the address of the signal level of the original image, Selector 5
5. A dual port as a gradation conversion table for reading out the signal level of the original image which is the output of No. 5 as an address
The random access memory includes a digital signal processor 71, a histogram storage memory 72a, a read-only memory 73a, and a dual port random access memory.
The access memory 74 forms the histogram conversion unit 7 in FIG.

However, as will be described later in detail, the program stored in the read-only memory 73a is different from the conventional one.

Reference numeral 9 denotes a digital-to-analog converter for converting data read from the dual-port random access memory 74 into analog data.

A monitor 10 displays the output of the digital / analog converter 9.

Also in this case, the gradation of the analog / digital converter is 12 bits, and the gradation of the digital / analog converter is 8 bits.
Bits and display pixels are assumed to be 320 (horizontal) × 240 (vertical).

Therefore, the capacities of the first frame memory 52 and the second frame memory 53, the histogram creation memory 61, and the dual port random access memory 74 are the same as those of the first embodiment of the present invention shown in FIG. Is the same as the capacitance in each of the embodiments.

On the other hand, since the capacity of the histogram storage memory 72a stores n (n is a positive integer) frames of histograms in the histogram storage memory, as will be described in detail later, the capacity is 17 bits × 12 bits × There are n frames.

The creation of a histogram is performed as follows.

First, the creation of the histogram of the original image is performed by the histogram creation memory 61 and the adder 62, using exactly the same technique as that already described.

Since the histogram of the original image is created sequentially, the creation of the histogram of the original image and the writing of the original signal level to the frame memory are performed simultaneously.

In the histogram creation memory 61, 1
When the histogram of the original image is created for the images in the frame memory, the created histogram of the original image is transferred to and stored in the histogram storage memory 72a.

The histogram of the original image is continuously transferred and stored over n frames from the histogram creation memory 61 to the histogram storage memory 72a.

When the histogram of the original image for n frames is stored in the histogram storage memory 72a, the digital signal processor 71 writes the histogram into the histogram storage memory 72a according to the program stored in the read-only memory 73a. The histograms of the original images for n frames are added and divided by the number n of added frames to obtain a histogram which is the average of the histograms of the original images over n frames.

Further, the digital signal processor 71 has a function of n
With respect to the average value of the histogram of the original image over the frame, the signal level where the number of pixels is equal to or less than the threshold value is cut, and the remaining signal level number N is equally allocated to the 8-bit signal level, and the allocated 8 bits Is written to the dual port random access memory 74 using the signal level of the original image as an address.

[0177] Dual port random access
The 8-bit gradation data written in the memory 74 is stored in the first frame memory 52 or the second frame memory 5.
3 is read out as an address using the signal level of the original image read out from any one of the monitors 3 and supplied to the digital-to-analog converter 9 and converted into an analog signal.
0 is supplied and displayed.

In this case, gradation conversion is performed on the average of n frames of histograms stored in the histogram storage memory. That is, the histogram subjected to the gradation conversion is not a histogram of a specific frame.

In the second embodiment of the present invention, it is added that processing is performed uniformly for all temperature ranges instead of obtaining and converting histograms for various temperature ranges. deep.

FIG. 7 is a diagram illustrating the above procedure as a diagram illustrating a processing procedure according to the second embodiment of the present invention.

FIG. 7A shows a frame.

FIG. 7B shows valid data of the original image data.

FIG. 7C shows the storage of valid data of the original image in the frame memory. The acquired valid data is sequentially and alternately stored in the first and second frame memories for each frame.

At the same time that the effective data is stored in the frame memory, the signal level of the same effective data is supplied to the histogram creation memory, and as described above, the histogram of the original image is successively created during one frame. It is being done. This is shown in FIG.

1 created by the histogram creation memory
The histogram of the original image for the frame is transferred to the histogram storage memory during the idle time of the same frame. When the histograms of the n frames of the original image are stored, the addition operation and the division by n are performed and averaged at the beginning of the frame immediately after the last frame in which the histogram is stored (FIG. 7E). )), The above-described conversion of the histogram is performed during the remaining time of the same frame (FIG. 7F), and the result is stored in the dual port random access memory 74.

[0186] Dual port random access
The conversion result of the histogram stored in the memory 74 is obtained by using, as an address, the signal level of the original image written in the first frame memory or the second frame memory in the frame immediately before the frame in which the conversion result was obtained. Is read. With this, the signal level after gradation conversion is digital
It will be supplied to the analog converter. This is shown in FIG.
This is shown in “output data (n + 2)” of (g).

In FIG. 7, the processing procedure is illustrated by focusing on the valid data and the histogram of a specific range of frames in order to avoid complication, but the above operation is always continuous. Needless to say, this is being done.

In the configuration shown in FIG. 1, a frame memory needs a capacity of three frames in order to perform gradation conversion by associating the original image of the same frame with the gradation-converted histogram. However, in the configuration of FIG.
The gradation conversion is performed by associating the histogram obtained by averaging the histogram of the frame with the original image. That is, the signal level subjected to the gradation conversion is equal to the signal level of the original image of the specific frame by 1:
Since the relationship is not the same, in principle, the frame memory may be composed of two random access memories.

FIG. 8 is a view for explaining image processing according to the second embodiment of the present invention.

In FIG. 8, the vertical direction is the time axis, and the lower the figure, the more time elapses.

The horizontal direction of FIG. 8 shows an original image, a histogram of the original image, a histogram after averaging, and a histogram after gradation conversion.

Assuming that only the background appears in the original image for a long time, the histogram of the original image and the averaged histogram are the same at first. Since the 12-bit signal level of the averaged histogram at this time is equally assigned to the 8-bit signal level, the entire 8-bit signal level is distributed to the background portion in the histogram after gradation conversion. I have.

At the next time, a high-temperature object is assumed to appear in the original image. At this time, the high-temperature range of the histogram of the original image
As shown in the histogram of the original image in the second row, a peak of the histogram indicating the high-temperature object appears.

The averaging over n frames is performed over the histogram of the original image and the histogram of the original image of the previous (n-1) frame. Therefore, 2 rows 2
In the average of the histogram of the original image including the histogram of the original image in the column, small peaks appear in the high temperature region as shown in the averaged histogram in the second row and the third column. Since the gradation conversion is performed including the small peaks in the high-temperature region, the histogram after the gradation conversion has small peaks in the high-temperature region as shown in the second row and the fourth column. Slightly shifts to the left.

When a certain period of time has passed since such processing is repeated and the high-temperature object comes to appear in the original image, the histogram of the original image and the histogram after averaging become equal as shown in the bottom row of FIG. Accordingly, in the histogram after the gradation conversion, the peak indicating the histogram in the high-temperature region becomes larger, and the peak of the histogram indicating the background shifts further to the left from the beginning.

However, the speed at which the background mountain shifts to the left can be reduced by averaging. Therefore, according to the second embodiment of the present invention as well, it is possible to prevent a decrease in the visibility of an image when a high-temperature object moves at high speed in an original image. And naturally, even if a low-temperature object moves or a low-temperature object and a high-temperature object move, it can prevent that the visibility of an image falls.

FIG. 9 is a flowchart showing the operation of the second embodiment of the present invention. Hereafter, in a summary sense,
The operation of the second embodiment of the present invention will be described along the reference numerals in FIG.

S21. The histogram creation memory 61 in FIG. 6 is reset to create a state in which 0 is stored in all addresses.

This is an indispensable process at the first stage of the histogram creation.

S22. 1 is assigned to the address corresponding to the output of the analog-to-digital converter in the histogram creation memory.
Is added and stored in the corresponding address of the histogram creation memory.

As a result, a histogram is sequentially created for each signal level output from the analog-to-digital converter.

S23. It is determined whether the operation in step S22 has been completed for all pixels.

Since the number of pixels in one frame is known, it is only necessary to determine whether or not the count value of the counter which advances by the increment has reached the above-mentioned number of pixels each time it is added.

If it is determined that the processing has not been completed for all the pixels (No), the process jumps to step S22 and repeats the processing from step 22 to step S23 until the above operation is completed for all the pixels. Do.

The above is the stage of creating the histogram of the original image, the histogram creation memory 61 and the adder 62 shown in FIG.
Done by

S24. When it is determined in step S23 that the addition operation has been completed for all pixels (Yes)
, The histogram of the original image created in the histogram storage memory 72a of FIG. 6 is transferred from the histogram creation memory.

S25. It is determined whether the histogram of the original image for n frames has been transferred.

If it is determined that the histogram of the transferred original image is less than n frames (No), the process jumps to step S21 and repeats the processes of steps S21 to S24.

S26. If it is determined in step S25 that the histograms of the n frames of the original images have been transferred, the histograms of the n frames of the original images are added.

It should be noted that the transfer of the histograms of the n frames of the original image can be understood by monitoring the completion of the access to the n-th frame of the histogram storage memory 72a in FIG.

Once the access to the n-th frame of the histogram storage memory 72a in FIG. 6 is completed, the writing is performed cyclically, so that the original image for n frames is shifted by one frame at a time. Is stored, so that step S
In 25, it is always determined that n frames have been stored.

S27. The histogram added in step S26 is divided by the number n of the added frames to calculate a histogram obtained by averaging the histogram of the original image.

Steps S24 to S27 are steps of averaging the histogram, and are mainly performed by the digital signal processor.

S28. Signal level of original image 0-4095
, The signal level below a predetermined number of pixels is cut.

In this case, when the number of gradations is compressed and converted, a signal level that is not necessary to be displayed because of a low frequency of appearance is deleted, and an image is displayed efficiently within the number of compressed gradations. This is the process to be performed.

S29. Rearrange the remaining signal levels,
The total number N of levels is obtained.

S30. A straight line 4 having a value ranging from 0 to 1 with a slope of 1 / N is defined.

If i is an integer from 0 to N and the equation of the above straight line is L ₄ , then L ₄ = i (1 / N).

S31. The display level is converted to 8 bits by multiplying the straight line obtained in step S30 by the level 255 after gradation conversion.

That is, the display level after conversion to 8 bits is: display level = 255 · L ₄ .

S32. The display level obtained in step S31 is changed to the dual port random access
It is stored in the memory at an address corresponding to the signal level of the original image.

Steps S28 to S32 are steps of converting the histogram, and are mainly performed by the digital signal processor.

S33.2 The signal level of the frame memory to which the signal level of the original image is not written out of the frame memory of the plane 2 is stored as an address in the dual port random access memory. The read grayscale data is read and grayscale conversion is performed.

S34. The data read in step S33 is output to a digital-to-analog converter for analog conversion, and output to a monitor for image display.

Steps S33 and S34 are steps of gradation conversion and display, and gradation conversion is mainly performed by a digital signal processor.

FIG. 6 shows an example in which a histogram storage memory is provided for n frames and addition and averaging are performed by a digital signal processor. However, cumulative addition is performed by a known random access memory and an adder. The averaging can be performed by a circuit and the result of addition is divided by a divider, that is, the averaging can be entirely performed by hardware.

As another modification of the second embodiment,
Although not shown, in each of the two-sided frame memories, the signal levels of n frames are averaged in synchronization with the averaging of the histograms of the n frames, so that dual
The address may be a read address from the port random access memory.

With the above-described configuration of the two-frame memory, the signal level of the image read from one of the two-frame memories when performing the gradation conversion,
Since a reliable correlation is maintained between the signal levels in the averaged histogram, the image quality after gradation conversion is improved.

Next, the description moves to the description of the third embodiment of the present invention.

FIG. 10 shows a hardware configuration according to the third embodiment of the present invention.

In FIG. 10, reference numeral 1 denotes an optical system which collects and scans infrared light to be received and 2 denotes an infrared detector which receives and scans the infrared light scanned by the optical system 1 and outputs an analog electric signal; Reference numeral 3 denotes an amplifier for amplifying the output of the infrared detector 2 to a predetermined level in order to secure a signal-to-noise ratio of the analog electric signal output from the infrared detector 2.
Is an analog / digital converter for converting the analog output of the amplifier 3 into a digital signal.

Reference numeral 51a denotes a third selector for sequentially outputting the output of the analog-to-digital converter 4 to different output terminals.
52 is a first frame memory for storing the signal level of the original image for one frame, 53 is a second frame memory for storing the signal level of the original image for one frame, 55a
Is the first frame memory 52 or the second frame memory.
A fourth selector for selecting one of the outputs of the memory 53,
A third selector 51a, a first frame memory 52,
Second frame memory 53 and fourth selector 55a
The frame memory 5 shown in FIG.

The signal levels of the original image are stored in the first frame memory 52 and the second frame memory 53 using the pixel address of the original image as an address. From the frame memory 53, the signal level of the original image is read using the pixel address of the original image as an address.

Reference numeral 61 denotes a histogram creation memory for creating a histogram of an original image; 62, an adder for adding 1 to the output of the histogram creation memory 61; 63, a buffer gate;
1, the adder 62 and the buffer gate 63 constitute the histogram creating unit 6 of FIG.

Reference numeral 81 denotes a first weighting calculation memory for storing a histogram obtained by performing a weighting calculation described later on the histogram created by the histogram creation memory 61, and reference numeral 82 denotes a histogram created by the histogram creation memory 61. A second weighting operation memory 83 for storing a histogram subjected to a weighting operation to be described later, 83
Is subtraction between the histogram of the original image created by the histogram creation memory 61 and the histogram weighted and calculated one frame before stored in the first weighting calculation memory or the second weighting calculation memory 82. Is a read-only memory that stores a weighting coefficient for the signal level of the original image using the signal level of the original image as an address, and 85 is a weighting coefficient read from the read-only memory 84 and the output of the adder 83. The multiplier 86 multiplies the output of the multiplier 85 by:
The histogram read out from the first weighting calculation memory 81 or the second weighting calculation memory 82 and subjected to the weighting calculation one frame before is added, and the addition result is added to the first weighting calculation memory 81 or the second weighting calculation memory. The adder 87 that supplies the weighting calculation memory 82 of the
The fifth selector 88 that supplies the output of the first weighting operation to one of the first weighting operation memory 81 and the second weighting operation memory 82 includes a first weighting operation memory 81 or a second weighting operation memory 82 A first weighting calculation memory 81, a second weighting calculation memory 82, an adder 83, a read-only memory 84, a multiplier 8
5, the adder 86, the fifth selector 87, and the sixth selector 88 constitute a weighting operation unit.

Reference numeral 71 denotes a digital signal processor for performing histogram conversion, 72 denotes a histogram storage memory for storing the operation result of the weighting operation unit, and 73b denotes a histogram storage memory.
A read-only memory 74 storing a program to be executed by the digital signal processor 71 is a histogram obtained by the digital signal processor 71 converting a histogram from a histogram stored in the histogram storage memory 72 into a histogram.
A dual port random access memory as a gradation conversion table, wherein the signal level of the original image is written as an address and the signal level of the original image output from the fourth selector 55a is read as an address; 71, histogram storage memory 7
2. Read only memory 73b and dual port
The random access memory 74 constitutes the histogram converter 7 shown in FIG.

However, the program stored in the read-only memory 73b in the configuration of FIG. 10 is different from the conventional technology.

Reference numeral 9 denotes a digital-to-analog converter for converting image data read from the dual-port random access memory 74 into analog data.

Reference numeral 10 denotes a monitor for displaying the output of the digital / analog converter 9.

Also in this case, the gradation of the analog-to-digital converter is 12 bits, and the gradation of the digital-analog converter is 8 bits.
Bits and display pixels are assumed to be 320 (horizontal) × 240 (vertical).

Therefore, the capacities of the first frame memory 52, the second frame memory 53, the histogram creation memory 61, and the dual port random access memory 74 are the same as those of the first embodiment of the present invention shown in FIG. Is the same as the capacitance in the embodiment.

The operation of the histogram calculation unit is as follows.

The histogram creation memory 61, the first weighting calculation memory 81, and the second weighting calculation memory 82 are reset at the start of the operation.

After the histogram of the original image for one frame is output from the histogram creation memory 61, the difference from the storage content of the first weighting operation memory 81 or the second weighting operation memory 82 is obtained.

The difference is multiplied by a weight coefficient read from the read-only memory 84 using the signal level of the original image as an address. The weight coefficient is set to, for example, 1 for a normal temperature range of the signal level of the original image, and is set to a small value for a high temperature range. Therefore, in the output of the multiplier 85, the high temperature region is relatively compressed, and the normal temperature region is not compressed.

Next, the output of the multiplier 85 is added to the first weighting operation memory 81 or the second weighting operation memory 8.
2 are added. Initially, the values of the histograms stored in the first weighting operation memory 81 or the second weighting operation memory 82 are all 0, so that the addition result matches the output of the multiplier 85. This is stored in one of the first weighting operation memory 81 and the second weighting operation memory 82.

The same operation is repeated in the subsequent frames.

Therefore, the capacity of the first weighting operation memory 81 and the second weighting operation memory 82 required for performing the above operation is 17 bits × 12 bits.

In the third embodiment of the present invention, it is to be noted that the image processing is not performed by dividing the histogram of the original image into various temperature ranges, but by processing all the temperature ranges uniformly. Keep it.

FIG. 11 is a diagram illustrating the above procedure as a diagram showing a processing procedure according to the third embodiment of the present invention.

FIG. 11A shows a frame.

FIG. 11B shows valid data of the original image data.

FIG. 11C shows a frame of valid data.
This shows saving to memory. The acquired valid data is sequentially and alternately stored in the first and second frame memories for each frame.

At the same time that the signal level of the effective data is stored in the frame memory, the same effective data is supplied to the histogram creation memory, and the histogram is successively created during one frame as described above. . This is shown in FIG.

1 created in the histogram creation memory
The histogram of the original image for the frame is transferred to the histogram calculator during the idle time of the same frame. And
One frame before, stored in one of the first weighting calculation memory or the second weighting calculation memory,
The weighting operation with the stored weighted histogram is performed, and the histogram is again stored in a weighting operation memory different from the weighting operation memory that stores the histogram used in the operation (FIG. 11 (e). )) Is transferred to the histogram storage memory.

As described above, the histogram of the original image for one frame created by the histogram creation memory and the one frame stored in one of the first weighting operation memory and the second weighting operation memory Since the weighting calculation with the previously calculated and stored weighted histogram is performed and the histogram used for the calculation is stored again in a different weighting calculation memory from the weighting calculation memory that stored the histogram, This means that two weighting memories are required.

The histogram conversion already described is performed on the histogram stored in the histogram storage memory (FIG. 11F), and the signal level of the original image is stored in the dual port random access memory. Stored as an address.

The histogram stored in the dual port random access memory is obtained by using the signal level of the original image written in the first frame memory or the second frame memory one frame before as an address. Is read. Thus, the signal level subjected to the gradation conversion is supplied to the digital / analog converter. This is the “output data 4” in FIG.

Although FIG. 11 illustrates the processing procedure by focusing only on the valid data and the histogram of a specific frame in order to avoid the figure from being congested, the above operation is always performed continuously. Is being done.

In the configuration shown in FIG. 1, the frame memory needs a capacity of three frames in order to perform gradation conversion in correspondence with the original image and the histogram of the same frame. In the configuration, the difference between the histogram of the previous frame and the histogram are weighted, and the histogram obtained by performing the operation of adding the weighted difference to the histogram of the previous frame and the original image are associated with each other to perform the gradation conversion. Therefore, the histogram at the time of gradation conversion is not for a specific frame. Therefore, in principle,
The frame memory only needs to have a capacity for two frames.

FIG. 12 is a view for explaining image processing according to the third embodiment of the present invention.

In FIG. 12, the vertical direction is the time axis, which means that the time elapses toward the bottom of the figure.

In the horizontal direction of FIG. 12, an original image, a histogram of the original image, a histogram after weighting, and a histogram after gradation conversion are shown.

Assuming that only the background appears in the original image for a long time, the histogram of the original image and the histogram after weighting are the same at first. Since the 12-bit signal level of the weighted histogram at this time is equally assigned to the 8-bit signal level, the signal level is distributed to the background over the entire 8 bits in the histogram after gradation conversion.

At the next time, the high-temperature object is assumed to appear in the original image. At this time, the high-temperature range of the histogram of the original image
As shown in the histogram of the original image in the second row, a peak of the histogram indicating the high-temperature object appears.

The weighting for the previous frame is performed between the histogram of the original image and the weighted histogram stored in the weighting calculation memory one frame before. That is, first, the difference between the histogram of the original image in the second row and the second column and the histogram of the first row and the third column is calculated. In this case, since the histogram of the background normal temperature region is equal, the difference is equal to the peak of the high temperature region in the histogram of the original image in the second row and second column.

As described above, since a small weighting factor is applied to the temperature information in the high temperature range, the histogram of the difference after weighting has a small peak in the high temperature range. Since the difference after the weighting and the histogram stored in the weighting operation memory one frame before are added, the result is as shown in the histogram after the weighting in the second row and the third column in FIG.

When a certain period of time has passed since such processing is repeated and the high-temperature object comes to appear in the original image, the histogram of the original image and the histogram after weighting are equal as shown in the bottom row of FIG. Accordingly, in the histogram after the gradation conversion, the peak of the histogram indicating the high-temperature region becomes large, and the peak of the histogram indicating the background shifts to the left from the beginning.

However, the speed at which the background mountain shifts to the left can be reduced by weighting. Therefore, according to the third embodiment of the present invention, the high-temperature object moves at high speed in the original image. In this case, it is possible to prevent the visibility of the image from lowering. Of course, even if the low-temperature object moves at a high speed in the original image, or if the low-temperature object and the high-temperature object move at a high speed, it is possible to prevent the visibility of the image from lowering.

FIG. 13 is a flowchart showing the operation of the third embodiment of the present invention. Hereinafter, the operation of the second embodiment of the present invention will be described along the reference numerals in FIG.

S41. The histogram creation memory 61 and the weighting calculation memory 91 in FIG. 10 are reset to create a state in which 0 is stored in all addresses.

This is an essential process in the first stage of histogram creation.

S42. 1 is assigned to the address corresponding to the output of the analog-to-digital converter in the histogram creation memory.
Is added and stored in the corresponding address of the histogram creation memory.

As a result, a histogram of the original image is sequentially created for each signal level output from the analog-to-digital converter.

S43. It is determined whether the operation in step S42 has been completed for all pixels.

Since the number of pixels in one frame is known, it is only necessary to judge whether or not the count value of the counter which advances by the increment has reached the above-mentioned number of pixels each time the addition is performed.

If it is determined that the processing has not been completed for all the pixels (No), the process jumps to step S42 and repeats the processing from step S42 to step S43 until the above operation is completed for all the pixels. Do.

The above is the stage of creating the histogram of the original image, and the histogram creation memory 61 and the adder 6 shown in FIG.
2 is performed.

S44. When it is determined in step S43 that the addition operation has been completed for all pixels (Yes)
, The histogram of the original image created in the weighting calculation unit in FIG. 10 is transferred from the histogram creation memory.

S45. The difference between the transferred histogram of the original image and the weighted histogram stored in one of the weight calculation memories is calculated.

S46. The difference is multiplied by a different weighting coefficient for each temperature range.

The weighting is large for the normal temperature range in the background (normally, it is sufficient to set it to 1), and small for the high temperature range.

S47. The histogram calculated and stored one frame before and stored in the weight calculation memory is added to the weighted difference.

S48. The histogram of the addition result is
The data is transferred to the histogram storage memory and stored in the other weighting calculation memory.

Thereafter, this calculation is sequentially performed.

Then, steps S45 to S48 are the stages of the weighting calculation, and are performed by the weighting calculation unit of FIG.

S49. Signal level of original image 0-4095
, The signal level below a predetermined number of pixels is cut.

This is because, when the gradation number is compressed and converted, a signal level having a low appearance frequency and a low display necessity is deleted, and an image is displayed efficiently within the compressed gradation number. This is the process to be performed.

S50. Rearrange the remaining signal levels,
The total number N of signal levels is obtained.

S51. A straight line 4 having a value ranging from 0 to 1 with a slope of 1 / N is defined.

If i is an integer up to 0 and N and the equation of the straight line is L ₄ , then L ₄ = i (1 / N).

S52. The display level is converted into 8 bits by multiplying the straight line obtained in step S51 by the level 255 after gradation conversion.

That is, the display level after conversion to 8 bits is: display level = 255 · L ₄ .

Steps S49 to S52 are steps of histogram conversion, and are mainly performed by the digital signal processor.

S53. The display level obtained in step S52 is converted to a demographic port
It is stored in an address corresponding to the signal level of the original image in the random access memory.

S54.2 The signal level of the frame memory to which the signal level of the original image is not written out of the frame memory of the surface is stored in the dual port random access memory as an address. The gradation data is read and gradation conversion is performed.

S55. The data read in step S33 is output to a digital-to-analog converter for analog conversion, and output to a monitor for image display.

Steps S54 and S55 are steps of gradation conversion and display, and gradation conversion is mainly performed by a digital signal processor.

Although FIG. 10 shows an example in which the weighting operation for the histogram of the original image is performed by hardware called a weighting operation unit, the weighting operation can be performed by software using a digital signal processor.

As another modification of the third embodiment of the present invention, although not shown, the analog-to-digital converter 4
In a two-sided frame memory that stores the signal levels of the original images output by the above, it is also possible to perform a weighting operation in synchronization with the above weighting operation to obtain a read address from the dual port random access memory. is there.

[0301] By employing the above-described configuration of the two-frame memory, the signal level of the image read out from one of the two-frame memories when performing the gradation conversion,
Since a reliable correlation is maintained between the signal levels in the averaged histogram, the image quality after gradation conversion is improved.

[0302] In the first to third embodiments of the present invention, the histogram conversion unit converts the input histogram or the histogram generated by the histogram conversion unit into a signal corresponding to a predetermined number of pixels or less. It is described as cutting the level. This is based on the practical determination that the same processing has been performed conventionally, and that it is not necessary to assign a signal level even after performing gradation conversion on a signal level with a low appearance frequency. It is not essential for image processing.

[0303]

According to the first means of the present invention, a histogram of an original image is created by dividing it into a plurality of temperature ranges, and gradation conversion is performed according to the number of signal levels in each of the plurality of divided temperature ranges. Since the number of subsequent signal levels is assigned, no change occurs in the number of signal levels assigned to the background portion after gradation conversion, even if there is, for example, a high-temperature object in the subject. Also, whether or not there is a low-temperature object in the subject does not change in the number of signal levels assigned to the background portion after the gradation conversion.

Therefore, even when the abnormal object moves within the subject at high speed, the visibility of the displayed image does not decrease.

According to the second means of the present invention, the histogram of the original image is averaged over a plurality of frames, and the signal level after gradation conversion is assigned, and the signal level of the original image and the signal level after gradation conversion are determined. Is performed, even if the abnormal object moves within the subject at high speed, the histogram of the background portion in the histogram after gradation conversion is not rapidly affected by the presence or absence of the abnormal object.

Therefore, even when the abnormal object moves at high speed in the subject, the visibility of the displayed image does not decrease.

According to the third means of the present invention, the difference between the histogram of the original image one frame before and the current histogram is obtained, a large weighting factor is applied to the normal temperature portion and the difference is applied to the high temperature portion, for example. For the histogram obtained by adding the weighted difference and the histogram of one frame before, adding the weighted difference and the histogram of one frame before, a signal level after gradation conversion is applied. Since the number is assigned, even if the high-temperature object moves within the subject at high speed, the histogram of the background portion in the histogram after gradation conversion is not rapidly affected by the presence or absence of the abnormal object. This is the same even if the low-temperature object moves at high speed in the subject.

Therefore, even when the abnormal object moves at high speed in the subject, the visibility of the displayed image does not decrease.

[Brief description of the drawings]

FIG. 1 is a hardware configuration according to a first embodiment of the present invention.

FIG. 2 is a diagram showing a processing procedure according to the first embodiment of the present invention.

FIG. 3 is a diagram illustrating the principle of the first embodiment of the present invention.

FIG. 4 is a view for explaining image processing according to the first embodiment of the present invention.

FIG. 5 is a flowchart showing the operation of the first embodiment of the present invention.

FIG. 6 is a hardware configuration according to a second embodiment of the present invention.

FIG. 7 is a diagram showing a processing procedure according to the second embodiment of the present invention.

FIG. 8 is a diagram illustrating image processing according to the second embodiment of the present invention.

FIG. 9 is a flowchart showing the operation of the second embodiment of the present invention.

FIG. 10 is a hardware configuration according to a third embodiment of the present invention.

FIG. 11 is a diagram showing a processing procedure according to the third embodiment of the present invention.

FIG. 12 is a diagram illustrating image processing according to a third embodiment of the present invention.

FIG. 13 is a flowchart showing the operation of the third embodiment of the present invention.

FIG. 14 is a general configuration of an infrared image display device.

FIG. 15 is a view for explaining conventional image processing.

FIG. 16 is a diagram illustrating a problem in conventional image processing.

[Explanation of symbols]

Reference Signs List 1 optical system 2 infrared detector 3 amplifier 4 analog / digital converter (A / D) 5 frame memory 6 histogram creation unit 7 histogram conversion unit 9 digital / analog converter 10 monitor 51 first selector (first SEL) ) 51a Third selector (third SEL) 52 First frame memory (RAM) 53 Second frame memory (RAM) 54 Third frame memory (RAM) 55 Second selector (second) 55a Fourth selector (fourth SEL) 61 Histogram creation memory (RAM) 62 Adder 63 Buffer gate 71 Digital signal processor (DSP) 72 Histogram storage memory (RAM) 73 Read-only memory (ROM) ) 74 Dual Port Random Access Memory (DP) RAM) 81 First weighting operation memory (RAM) 82 Second weighting operation memory (RAM) 83 Adder 84 Weighted read-only memory (ROM) 85 Multiplier 86 Adder 87 Fifth selector (Fifth selector SEL) 88 Sixth selector (Sixth SEL)

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) G06T 7/00 G06F 15/70 325 (72) Inventor Morito Shiobara 4-chome Kamiodanaka, Nakahara-ku, Kawasaki City, Kanagawa Prefecture No. 1 No. 1 F-term in Fujitsu Limited (reference) 2G066 BC07 BC09 BC30 CA08 5B057 BA08 CA08 CA12 CA16 CB08 CB12 CB16 CC01 CE03 CE11 CH11 DA16 DB02 DB09 DC23 DC32 5C082 AA12 BA16 BA27 BA35 BA39 BB14 BB15 BB26 CA10 A056 BA02 DA01 EA39 FA26 FA32 FA37 GA08 GA51 LA05

Claims (5)

[Claims] A histogram generation unit that divides a histogram, which is the frequency of appearance of a signal level in an original image, into a plurality of temperature regions; and generates a histogram output by the histogram generation unit. In accordance with the number of signal levels appearing in each of the above, the number of signal levels after gradation conversion is assigned to create a histogram after gradation conversion, and the signal level of the original image stored in the frame memory and the signal level after gradation conversion. An infrared image display device, comprising: 2. A histogram creating section for creating a histogram which is a frequency of appearance of a signal level in an original image by dividing the histogram into a plurality of temperature ranges, and averaging a histogram of the original image output by the histogram creating section over a plurality of frames. The signal level after gradation conversion is assigned to the histogram thus generated, and a histogram after gradation conversion is generated. The conversion between the signal level of the original image stored in the frame memory and the signal level after gradation conversion is performed. An infrared image display device, comprising: 3. The infrared image display device according to claim 2, wherein the frame memory averages image data of the original image over a plurality of frames in synchronization with the averaging of the histogram of the original image. -An infrared image display device characterized by being a memory. 4. A histogram creating section for creating a histogram which is a frequency of appearance of a signal level in an original image by dividing the histogram into a plurality of temperature ranges, and a histogram output by the histogram creating section and stored in a weighting calculation memory. The difference with the histogram is weighted by a different weighting coefficient for each temperature range, the weighted difference is added to the histogram output from the weight calculation memory, and the histogram creation memory outputs A weighting operation unit for performing a weighting operation on the histogram of the original image to be processed, and assigning the number of signal levels after the gradation conversion to the output of the weighting operation unit to generate a histogram after the gradation conversion,
An infrared image display device comprising: a histogram conversion unit that converts a signal level of an original image stored in a frame memory and a signal level after gradation conversion. 5. The infrared image display device according to claim 4, wherein the frame memory performs a weighting operation on the image data of the original image in synchronization with the weighting operation on the histogram of the original image. -An infrared image display device characterized by being a memory.