KR101041136B1 - Method for processing thermal image - Google Patents

Abstract

PURPOSE: A method for processing a thermal image is provided to provide a stable thermal image regardless of the temperature of an uncooled thermal observation device. CONSTITUTION: A method for processing a thermal image is as follows. Thermal image raw data, which constitute a frame by photographing an object through an uncooled thermal observation device, are obtained(S11). It is determined whether a difference between the temperature of an uncooled infrared rays detector and a pre-set reference temperature is existed or not(S12). The offset value which in advance compares with to the temperature of the ratio cooling thermal equipment when the photography is included to difference between the fixed reference temperature is produced. An offset value, proportional to the temperature difference, is calculated and is added to the thermal image raw data. The thermal raw data, in which the offset value is added, are converted into an offset correction value(S13).

Description

Method for processing thermal image

The present invention relates to a thermal image processing method, and to a method for performing real-time image processing without a TEC thermoelectric element in an uncooled thermal imager.

An infrared detector is a key component of thermal imaging equipment that detects the difference in inherent radiant energy emitted by targets and backgrounds at night when there is no light. The operating principle of the infrared detector detects infrared rays in the wavelength band of the infrared region emitted from the object (the long wave region has a wavelength of 8 to 12 µm and the medium wave region has a wavelength of 3 to 5 µm) and converts them into electrical signals. The electrical signal converted by the infrared detector as described above is displayed on the monitor, and appears as a constant image according to the temperature distribution of the object. Infrared detectors are variously classified according to the wavelength band and the cooling method. The detector can be classified into a non-cooling method used without cooling at room temperature and a cooling method cooled to cryogenic temperatures.

When operating an uncooled infrared detector, constant temperature maintenance of the infrared detector is essential for the acquisition and processing of the infrared signal. In order to maintain a constant temperature, an infrared detector requires a thermoelectric element called a TEC (Thermal Electrical Cooler) and a circuit for controlling the same.

That is, the thermal image output value of the uncooled infrared detector reacts differently according to the external temperature, so it must always be maintained at a constant temperature. For this purpose, a thermoelectric element called TEC is embedded in the hazardous detector. Implementing an uncooled infrared detector without a TEC thermoelectric device will result in noise and a fixed pattern. In addition, to maintain a constant temperature, the TEC thermoelectric element must be controlled with an error range of ± 0.01 °, which requires a separate control circuit design.

As a result, there is a problem of extra space for the TEC control circuit and power consumption for the TEC control. In addition, it may cause problems in terms of cost when mass production.

The technical problem of the present invention is to process a stable thermal image regardless of the temperature of the uncooled thermal imaging equipment. It is also an object of the present invention to provide an algorithm for removing noise and fixed patterns in uncooled thermal imaging equipment without TEC thermoelectric elements. In addition, the technical problem of the present invention is to provide a non-cooled thermal imaging equipment without a TEC thermoelectric element, thereby reducing the manufacturing cost and miniaturization of the non-cooling thermal imaging equipment.

An embodiment of the present invention provides a low data acquisition process of acquiring a thermal image raw data forming one frame by photographing a subject through an uncooled thermal imager, and a temperature and a preset reference temperature of the uncooled thermal imager when the image is taken. An offset correction value conversion process of converting the thermal image raw data into an offset correction value and considering whether a difference between the temperature of the uncooled thermal imaging apparatus and a reference temperature occurs, the thermal image raw data or offset After the correction value is selected, the image conversion process includes an image output process.

The offset correction value conversion process is performed through various interpolation methods of converting the thermal image raw data into an offset correction value.

 The image output process may be performed by performing an image conversion process on the offset correction value when the temperature of the uncooled thermal imaging equipment is smaller or larger than a preset reference temperature, and by performing an image conversion process on the thermal image raw data. Output

According to the embodiment of the present invention, a stable thermal image without noise may be provided regardless of the temperature of the uncooled thermal imager. In addition, according to the embodiment of the present invention, it is possible to manufacture a small size uncooled thermal imaging equipment without a TEC thermoelectric element. Further, according to the embodiment of the present invention, a low power uncooled thermal imager can be manufactured. Moreover, according to embodiment of this invention, manufacturing cost unit cost can be reduced.

1 is a flowchart illustrating a thermal image processing process in an uncooled infrared detector according to an exemplary embodiment of the present invention.
FIG. 2 is a thermal image of one [X, Y] pixel in a captured frame after photographing the same target subject by maintaining a different temperature of an uncooled infrared detector without a TEC thermoelectric device according to an exemplary embodiment of the present invention. This graph shows the output value.
3 is a graph illustrating thermal image raw data for each temperature of every pixel in an image frame according to an exemplary embodiment of the present invention.
4 is a graph showing a constant offset correction value in all temperature bands in one pixel according to an embodiment of the present invention.
5 is a graph showing a constant offset correction value irrespective of temperature for all pixels according to an exemplary embodiment of the present invention.
6 is a photograph showing an example of NUC image processing in an uncooled infrared detector without a TEC thermoelectric element.

Hereinafter, with reference to the accompanying drawings will be described an embodiment of the present invention in more detail. However, the present invention is not limited to the embodiments disclosed below, but will be implemented in various forms, and only the embodiments are intended to complete the disclosure of the present invention and to those skilled in the art to fully understand the scope of the invention. It is provided to inform you. Like numbers refer to like elements in the figures.

1 is a flowchart illustrating a thermal image processing process in an uncooled infrared detector according to an exemplary embodiment of the present invention.

In the following description, an infrared detector will be described as an example of an uncooled thermal imager that outputs a thermal image. However, it will be apparent that the embodiment of the present invention is not limited to an uncooled infrared detector but may be applied to all uncooled thermal imaging equipment capable of outputting a thermal image.

An uncooled infrared detector detects the infrared rays of an object (subject) and converts them into electrical signals.The principle of operation is that the wavelength band of the infrared region emitted by the object (long-wave region is 8-12㎛, medium-wave region is 3-5) Infrared rays having a wavelength of 占 퐉 are detected and data (hereinafter, referred to as "thermal image raw data") obtained by converting them as electrical signals is acquired in units of frames (S11).

As described above, the uncooled infrared detector acquires thermal image raw data in the form of an electrical signal by detecting infrared rays of the subject, and the obtained thermal image raw data is digitized to a predetermined bit amount. In this case, the thermal image raw data acquired from a subject of the same target may not have a constant value but have a constant value according to the temperature of the uncooled infrared detector.

This is because the uncooled infrared detector without the TEC thermoelectric element cannot maintain constant temperature, so the temperature of the uncooled infrared detector also changes as the external temperature changes, which results in a constant thermal image output of the thermal image for the same target. Because it does not appear. In a state in which thermal image raw data, which is a thermal image output value, does not appear constantly and has a value that varies with temperature, noise and a fixed pattern are generated when image processing is completed using the thermal image raw data.

 For reference, FIG. 2 shows one [X, Y] pixel in the captured frame, for example, [1,1] after photographing the same target subject by maintaining the temperature of the uncooled infrared detector without the TEC thermoelectric element differently. A graph showing thermal image output values (thermal image raw data) for pixels.

The temperature change of the uncooled infrared detector was made in the range of -10 ℃ ~ 10 ℃, and the thermal image low data, which is the thermal image output value obtained at this time, sets the value of Min 1 [V] ~ Max 4 [V] to 14 bits (0). ~ 16383).

2, it can be seen that the higher the uncooled infrared detector temperature, the higher the output value for the same temperature scene. In addition, since the value of the thermal image raw data increases as a slope of a predetermined pattern, it is possible to predict the value of the thermal image raw data at a specific temperature.

While FIG. 2 is a diagram illustrating temperature-specific thermal image raw data for one pixel in an image frame, for reference, FIG. 3 illustrates temperature-specific thermal image raw data for all respective pixels in an image frame. One graph. Referring to FIG. 3, each color line having a different color is a diagram illustrating a change in temperature of thermal image raw data of each pixel existing in a frame. If it is assumed that 100 pixels of 10 * 10 pixels are included in one frame, the color lines of FIG. 3 may include 100 color lines to represent values of thermal image raw data according to temperature change for each pixel.

Meanwhile, after the uncooled infrared detector acquires the raw image data of the frame photographed at the current temperature (S11), it is determined whether the temperature of the uncooled infrared detector when the photographing is made is different from the preset reference temperature (S12). ).

Thereafter, in consideration of the difference between the temperature of the non-cooled thermal imaging equipment and the preset reference temperature when photographing is performed (S12), an offset correction value conversion process of converting the thermal image raw data into an offset correction value is performed (S13). ).

The offset correction may be corrected through various interpolations such as linear and nonlinear. The linear interpolation method may be performed between a temperature of an uncooled thermal imager and a preset reference temperature when the imaging is performed. In proportion to the difference, the thermal image raw data is converted into an offset correction value. An example of linear interpolation will be briefly described.

Referring to FIG. 2, the [1,1] pixel has a thermal image low data value of '8500' at −10 ° C., and gradually changes as the temperature increases, and thus the thermal image low data value of “8750” at 0 ° C. It can be seen that has a thermal image low data value of '9000' at 10 ℃.

Therefore, it can be seen that the [1,1] pixel has a linear slope of 25 (data / ° C.) according to the temperature change. The offset correction is made in consideration of the change of the linear slope.

For example, it is assumed that the temperature of the uncooled infrared detector at the time of photographing is -10 ° C, and a thermal image low data value of '8500' is obtained. When the reference temperature is 0 ° C., since the reference temperature is 10 ° C. higher than the temperature of the uncooled infrared detector at which imaging is performed (S12), when linear interpolation is performed in consideration of this, that is, 25 (data / ° C.) × An offset value of 10 ° C. = 250 can be obtained. The offset value 250 may be added to a thermal image low data value of 8500 obtained by photographing at a temperature of an uncooled infrared detector of −10 ° C. to obtain an offset correction value of “8750” (S13).

As a result, when there is a difference between the temperature of the uncooled infrared detector and the reference temperature, the currently obtained thermal image low data may be offset-corrected to the thermal image low data value at the reference temperature. Therefore, if the correction is applied to the offset value, a constant offset correction value can be obtained as shown in FIG. 4 in all temperature bands.

On the other hand, each pixel in the frame has a different offset value, because the inclination of the value of the thermal image raw data for each temperature is different for each pixel as shown in FIG. By applying an offset value different for each pixel to offset correction for each pixel, a constant offset correction value irrespective of temperature can be obtained for all pixels as shown in FIG. 5.

For reference, Table 1 below is a table showing an example of offset values of each pixel, assuming that one frame has 10 x 10 pixels.

pixel Offset value [1,1] 25 [1,2] 26 ...
...
... ...
...
... [10,8] 34 [10,9] 36 [10,10] 39

When the offset correction is made as described above, the image conversion processing is performed with the offset corrected offset correction value and output through the display means such as a display (S14).

That is, according to whether a difference between the temperature of the uncooled thermal imaging equipment and the reference temperature occurs, the thermal image raw data or the offset correction value is selected and then image converted to output. For example, when the temperature of the uncooled thermal imaging equipment is smaller or larger than a preset reference temperature, the image is subjected to image conversion processing for the offset correction value, and if the same, the image conversion processing for the thermal image raw data is output. Have

The processes S11, S12, S13, and S14 are repeatedly performed for each frame until reaching the last frame (S15).

On the other hand, the process of performing the image conversion process (S14) may be made by a conventionally known image conversion processing method, for example, the image by the non-uniform correction (NUC) and contrast enhancement (CEM; Contras Enhancement Mapping) Treatment can take place.

Nonuniformity Correction (NUC) improves the uniformity of image quality by correcting gain and offset differences to make the output size constant for the same input signal. The thermal image quality depends on the detector non-uniformity correction. The non-uniformity correction reduces the spatial noise of the image to ensure maximum thermal performance and maximum system performance. Through image processing of non-uniformity correction, the values of all pixels can be converged into one feature value.

The general equation of the non-uniformity correction image processing is as shown in [Equation 1].

[Equation 1]

Y = G × RAW + NUC_Offset

In the above description, Y is a non-uniformly corrected (NUC) image-processed output value, G is signal gain, RAW is thermal image raw data, and NUC_Offset is an offset value during NUC image processing, and is different for each pixel.

In the embodiment of the present invention, the offset correction value (offset) obtained in step S13 is applied in the form of 'RAW + offset' instead of 'RAW', and thus, may be represented by the following [Equation 2].

[Equation 2]

Y = G × (RAW + offset) + NUC_Offset

Contrast Ratio Enhancement (CEM) image processing is to improve and improve the contrast ratio of an image in order to improve the discriminating power of an object. In the case of a thermal image, the concentration distribution is biased in some regions, and thus, the image having a low contrast ratio may be considered. Therefore, if the contrast ratio is low, the image is not clear and the target identification becomes difficult. Therefore, the image quality is improved through the contrast enhancement image processing. A histogram smoothing method is used as a density conversion method commonly used in general contrast enhancement image processing.

For reference, FIG. 6 is a photograph showing an example of NUC image processing in an uncooled infrared detector without a TEC thermoelectric element. FIG. 6 shows noise and a fixed pattern when NUC image processing is performed by Equation 1 without offset correction due to temperature change. It can be seen that it occurs as 6 (a). However, after performing offset correction according to the temperature change according to the present invention, when NUC image processing is performed by Equation 2, a normal image from which noise and a fixed pattern are removed can be obtained as shown in FIG.

Although the invention has been described with reference to the accompanying drawings and the preferred embodiments described above, the invention is not limited thereto, but is defined by the claims that follow. Accordingly, one of ordinary skill in the art may variously modify and modify the present invention without departing from the spirit of the following claims.

S11: Acquire thermal image low data S12: Difference from reference temperature
S13: Convert to offset correction value S14: Perform image conversion processing
S15: Is it the last frame

Claims (7)

A row data acquisition process of acquiring a thermal image raw data constituting one frame by photographing a subject through an uncooled thermal imager;
After calculating an offset value that is proportional to the difference between the temperature of the uncooled thermal imaging equipment and the preset reference temperature when the image is taken, the thermal image raw data is added as an offset correction value by adding the offset value to the thermal image raw data. A process of converting offset correction values for conversion;
An image output process of selecting the thermal image raw data or the offset correction value and performing image conversion processing according to whether a difference between the temperature of the uncooled thermal imaging equipment and a reference temperature occurs;
Thermal image processing method comprising a. The method of claim 1, wherein the row data acquisition process, the offset correction value conversion process, and the image output process are performed frame by frame until the last frame is reached. delete The method of claim 1, wherein the offset correction value conversion process is performed only when there is a difference between a temperature of the uncooled thermal imaging equipment and a preset reference temperature when the photographing is performed. The image output process of claim 1, wherein the image output process outputs an image by performing an image conversion process on the offset correction value when the temperature of the uncooled thermal imaging equipment is less than or greater than a preset reference temperature. Thermal image processing method for outputting the image conversion processing. The method of claim 1, wherein the image conversion processing comprises performing image conversion processing of non-uniform correction (NUC) and contrast enhancement mapping (CEM). The non-uniformity correction method according to claim 6, wherein Y is an output value of non-uniformly corrected image processing, G is signal gain, RAW is thermal image raw data, NUC_Offset is an offset value during NUC image processing, and RAW + offset is an offset correction value. The thermal image processing method calculated by "Y = G x (RAW + offset) + NUC_Offset".

Images