KR101705063B1 - Defective pixel detection method and device using memorized position of thermal imaging camera - Google Patents

Abstract

The present invention relates to a defective pixel detection apparatus using a memorized position of a thermal imaging camera. The defective pixel detection apparatus using a memorized position of a thermal imaging camera includes: an infrared sensor for detecting infrared rays emitted from a subject to output thermal image data corresponding to the infrared rays; a storage unit for storing a non-uniformity correction (NUC) table having slope information using an output value of each pixel of the infrared sensor while changing a temperature of the subject, wherein, in a case of a defective pixel whose slope is out of a predetermined range, the storage unit stores a non-uniformity correction (NUC) table having a preset minimum or maximum value instead of the corresponding slope; a non-uniformity correction unit that corrects an output value of each pixel of the thermal image data output from the infrared sensor with reference to the uploaded NUC table when the NUC table stored in the storage unit is uploaded at boot time; and a defective pixel processing unit for searching the uploaded NUC table for each pixel slope value of the thermal image data, and correcting and outputting the value of the defective pixel if the value corresponds to the preset minimum or a maximum value, wherein the defective pixel processing unit outputs, without numerical correction, the slope value of each pixel of the thermal image data when the slope value does not correspond to the preset minimum or maximum value.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention [0001] The present invention relates to a defective pixel detection method and a defective pixel detection method using position memory of a thermal imaging camera,

The present invention relates to a defective pixel detection method and a defective pixel detection method using position memory of a thermal imaging camera, and more particularly, to a defective pixel detection method and a defective pixel detection method by recording the position of a defective pixel in a NUC (Non Uniformity Correction) And more particularly to a defective pixel detection method and a defective pixel detection method using a positional memory of a thermal imaging camera which does not require a separate memory for storing defective pixels and does not require a separate process for determining defective pixels.

A thermal imaging camera is a device which senses infrared rays radiated from an object and measures and visually displays the temperature. Such thermal imaging cameras are widely used in the industrial fields of electricity, machinery and equipment to identify objects at night.

In a thermal imaging camera, an image sensor is formed by arranging a plurality of pixels in a two-dimensional structure, and each pixel performs a function of differentiating an optical signal according to brightness and converting it into an electrical signal.

Generally, a defective pixel (DP) or a dead pixel (DP) that does not normally react to light may be generated due to various factors during the manufacturing process of the image sensor. The defective pixel may be a white dead pixel A black dead pixel (a pixel whose sensitivity is significantly lower than that of other pixels, generally detected on a high-contrast screen), a low-sensitivity pixel Reactive defective pixels (pixels whose reaction state is lower than adjacent pixels, usually having a reaction state difference of 10% or more).

1 shows an example in which a defective pixel is included in an image. Bad pixels will appear as white or black.

In the case of a high-resolution pixel (12M pixle: 4000x3000 pixel size) such as a mobile sensor, the algorithm can compensate. However, in the case of a thermal image camera having a high price, the pixel is low (80x60 pixel size) When compensation is made, the image to be displayed is often not displayed.

Even if only a few of these defective pixels occur in a few millions of pixels, the corresponding image sensor product must be discarded. Therefore, DPC (Defective Pixel Compensation), which detects defective pixels and compensates them through signal processing, is recognized as important.

Conventional bad pixel (DP) processing methods include the following.

First, it is a method of judging and compensating for defective pixels in real time. This method is mainly used in general image sensors. However, it is not an actual defective pixel due to an algorithm for determining a defective pixel, but it may be determined as a defective pixel, and a normal pixel may be compensated to have a different value.

Fig. 2 shows the pre-compensation image and the post-compensation image. As shown in the right upper part of FIG. 2, the actual heat is displayed, and the heating point can be eliminated by compensating for the defective pixel. That is, when it is applied to a fire monitoring system, there arises a problem that a fire occurred at a long distance can not be detected due to defective pixel compensation.

Secondly, the image is analyzed at the time of mass production, and the coordinates of the defective pixel are stored and compensated. An additional process for judging and storing defective pixels at the time of mass production is required, thereby increasing the mass production time. A programmable read-only memory (PROM) for storing coordinates of defective pixels is separately required. As a result, a separate memory for storing the position of the defective pixel as well as the additional process for determining the defective pixel is required.

Published Japanese Patent Application No. 10-2007-0098263

SUMMARY OF THE INVENTION The present invention has been made in order to solve the above problem, and it is an object of the present invention to provide a non-uniformity correction (NUC) table in which a characteristic curve of each pixel is obtained while changing the temperature of a subject during production, A defective pixel detection method which uses a positional memory of a thermal imaging camera which is stored in a predetermined minimum value or a maximum value instead of the obtained inclination value and which can be processed by using a nonuniform correction table and a nonuniform correction table uploaded from a memory during bad pixel processing, And to provide a detection device.

According to an aspect of the present invention, there is provided a method of controlling an infrared sensor, the method comprising: obtaining an inclination using an output value of each pixel of an infrared sensor while changing a temperature of a subject; Determining the pixel as a defective pixel if the slope is out of a predetermined range; And generating a non-uniformity correction (NUC) table by storing a slope for each pixel in a storage unit, and storing a non-uniformity correction (NUC) table as a preset minimum or maximum value instead of the obtained slope in the case of a defective pixel .

The step of generating the nonuniformity correction (NUC) table may store the predetermined minimum value instead of the obtained slope when the output value of the defective pixel is smaller than the output value of the normal pixel. If the output value of the defective pixel is larger than the output value of the normal pixel, As a predetermined maximum value.

The method may further include: uploading a non-uniformity correction (NUC) table generated by the method in the boot time storage unit; Detecting infrared rays emitted from the subject by the infrared sensor and outputting the detected infrared rays as thermal image data; Correcting an output value for each pixel of the column image data by referring to the uploaded NUC table; And the bad pixel processing unit searches the uploaded NUC table for the slope value of each column of the column image data, corrects the value of the corresponding bad pixel if it is the preset minimum value or the maximum value, When the slope value of each pixel does not correspond to a predetermined minimum value or maximum value, outputting without numerical correction.

Wherein the defective pixel processing unit corrects the numerical value to output values of adjacent pixels not corresponding to the minimum value and the maximum value when the slope value of each pixel of the thermal image data corresponds to a preset minimum value or maximum value.

And transmits the minimum value and the maximum value information stored in advance in the NUC table to the bad pixel processing unit after the nonuniformity correction (NUC) table is uploaded.

After the nonuniformity correction (NUC) table is uploaded, the bad pixel processing unit transmits information indicating that the minimum value and the maximum value information stored in advance in the NUC table are not received, or that there is no minimum value and maximum value stored in advance in the NUC table And outputs the read NUC table without performing numerical correction on each pixel of the column image data without searching the uploaded NUC table.

According to another aspect of the present invention, there is provided an infrared sensor comprising: an infrared sensor for detecting infrared rays emitted from a subject and outputting the infrared image data as thermal image data; A non-uniformity correction (NUC) table having slope information using an output value of each pixel of the infrared sensor while changing a temperature of a subject is stored. In case of a defective pixel whose slope is out of a predetermined range, A storage unit for storing a nonuniformity correction (NUC) table having a maximum value; A non-uniform correction unit for correcting an output value of each pixel of the thermal image data output from the infrared sensor by referring to the uploaded NUC table when the NUC table stored in the storage unit is uploaded at boot time; And searching the uploaded NUC table for the slope value of each column of the thermal image data, correcting and outputting the value of the corresponding bad pixel if it corresponds to a preset minimum or maximum value, And a defective pixel processing unit for outputting the value without performing the numerical correction when the value does not correspond to the predetermined minimum value or the maximum value.

The storage unit generates a non-uniformity correction (NUC) table with a predetermined minimum value instead of the obtained slope when the output value of the defective pixel is smaller than the output value of the normal pixel. If the output value of the defective pixel is larger than the output value of the normal pixel, And generates a nonuniformity correction (NUC) table with the set maximum value.

Wherein the defective pixel processing unit corrects the numerical value to output values of adjacent pixels not corresponding to the minimum value and the maximum value when the slope value of each pixel of the thermal image data corresponds to a preset minimum value or maximum value.

When the nonuniformity correction (NUC) table stored in the storage unit is uploaded at boot time, the control unit transfers the minimum value and the maximum value information previously stored in the NUC table to the bad pixel processing unit.

After the nonuniformity correction (NUC) table is uploaded, the bad pixel processing unit transmits information indicating that the minimum value and the maximum value information stored in advance in the NUC table are not received, or that there is no minimum value and maximum value stored in advance in the NUC table And outputs the read NUC table without performing numerical correction on each pixel of the column image data without searching the uploaded NUC table.

According to the present invention having the above-described configuration, the following effects can be achieved.

First, the present invention can solve the problem of having a different value while compensating for a defective pixel, though it is not an actual defective pixel by an algorithm for determining a defective pixel.

A separate additional memory for storing the position of the defective pixel DP is not required and a manufacturing process cost and a production time can be saved since a separate process for determining a defective pixel is not required for mass production.

In the mass production process, the position of the defective pixel is recorded at the time of generation of the nonuniformity correction (NUC) table and uploaded at the time of booting the system, so that the defective pixel processing section can quickly judge and process the defective pixel without a separate defective pixel determination algorithm have.

1 shows an example in which a defective pixel is included in an image.
Fig. 2 shows the pre-compensation image and the post-compensation image of the defective pixel (DP).
3 is a flowchart of a defective pixel detection method using position memory of a thermal imaging camera according to an embodiment of the present invention.
Fig. 4 shows characteristic curves for each pixel (pixel) obtained while changing the temperature of the subject. (a) shows a case of a normal pixel, and (b) shows a case of a defective pixel.
5 is a block diagram of a defective pixel detection apparatus that uses positional memory of a thermal imaging camera according to another embodiment of the present invention.
FIG. 6 shows an example in which a defective pixel is replaced with a preset minimum or maximum value when the NUC table is generated in the embodiment of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS The advantages and features of the present invention, and how to accomplish them, will become apparent by reference to the embodiments described in detail below with reference to the accompanying drawings.

However, the present invention is not limited to the embodiments disclosed below, but may be embodied in various forms.

The present embodiments are provided so that the disclosure of the present invention is thoroughly disclosed and that those skilled in the art will fully understand the scope of the present invention.

And the present invention is only defined by the scope of the claims.

Thus, in some embodiments, well known components, well known operations, and well-known techniques are not specifically described to avoid an undesirable interpretation of the present invention.

In addition, throughout the specification, like reference numerals refer to like elements, and the terms (mentioned) used herein are intended to illustrate the embodiments and not to limit the invention.

In this specification, the singular forms include plural forms unless the context clearly dictates otherwise, and the constituents and acts referred to as " comprising (or having) " do not exclude the presence or addition of one or more other constituents and actions .

Unless defined otherwise, all terms (including technical and scientific terms) used herein may be used in a sense commonly understood by one of ordinary skill in the art to which this invention belongs.

Also, commonly used predefined terms are not ideally or excessively interpreted unless they are defined.

[First Embodiment]

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

Referring to FIGS. 3 to 6, a defective pixel detection method using the positional memory of the thermal imaging camera according to the present invention proceeds in the following manner.

First, the slope is obtained by using the output value of each pixel of the infrared sensor 210 while changing the temperature of the subject (S10).

If the slope is out of the predetermined range, the pixel is determined to be a defective pixel (S20).

A non-uniformity correction (NUC) table 261 is stored in the storage unit 260 by storing a slope for each pixel, and in the case of a bad pixel, a non-uniformity correction (NUC) table (Step S30).

Specifically, when generating the non-uniformity correction (NUC) table 261, the output value of the defective pixel is stored as a predetermined minimum value instead of the obtained slope when the output value of the defective pixel is smaller than the output value of the normal pixel. Instead of the obtained slope, the maximum value is stored.

4, a characteristic curve is obtained for each pixel of the infrared sensor 210 while the temperature of the subject is changed to High, Middle, and Low. The slope of each pixel is the same concept as the gain used in NUC correction. A nonuniformity correction (NUC) table is generated using the characteristic value (slope) of each pixel. (a) Normal pixels show similar characteristics, but (b) Bad pixels show lower or upper biased characteristics. Since the normal pixel is obtained in a region having a similar gain value, the gain value of the defective pixel is set to the minimum value or the maximum value and used as the coordinates of the defective pixel. Referring to FIG. 6, a more detailed description will be given below.

In the product mass production process, the nonuniformity correction (NUC) table 261 is generated and stored in the storage unit 260 through the above procedure, and then the following process is performed.

At the boot time storage unit 260, the nonuniformity correction (NUC) table 261 generated by the above methods is uploaded (S40). Conventionally, a non-uniformity correction table and a position value of a defective pixel are uploaded to each memory, respectively. However, according to the present invention, only one non-uniformity correction table 261 can be uploaded from the memory. This is because the position of the defective pixel is stored in the nonuniformity correction table 261.

The infrared sensor 210 detects infrared rays emitted from the subject and outputs the infrared image data as thermal image data (S50). The preprocessing unit 220 may perform processing for correcting a defect of each pixel of the output row data.

The non-uniformity correction unit 230 refers to the uploaded NUC table 261 and corrects the output value of each pixel of the column image data (S60). Non-uniformity correction refers to correcting different gains for each pixel to a normalized value, and outputting the same output under the same conditions.

The defective pixel processing unit 240 searches the uploaded NUC table 261 for the slope value of each column of the column image data, and when it corresponds to the preset minimum or maximum value (S70), the defective pixel processing unit 240 corrects the value of the defective pixel (S80). If the slope value of each pixel of the column image data does not correspond to the preset minimum or maximum value (that is, when the slope value is a normalizer) (S90). The post-processing unit 250 performs color adjustment processing and the like on the thermal image data output from the defective pixel processing unit 240 and transfers the color image to the display side.

This process is repeated until the end of the image size is reached (S100). A method of numerically correcting defective pixels and outputting them is well known to those skilled in the art, so a detailed description thereof will be omitted.

6 shows an example of the NUC table. As described above with reference to FIG. 4, in the case of a normal pixel, the Gain value is not significantly different.

For example, when the gain value is 8 bits, the normal pixel gain value has a value near 0x80. If the gain value deviates greatly from the normal range, set the gain value to the minimum value or the maximum value. If 0x40 is set to the normal range before and after 0x80, pixels not corresponding to this range are determined to be defective pixels. As a result, the gain value of the defective pixel DP may be 0 to 0x3F or 0xC0 to 0xFF.

6 shows an example in which Ox00 corresponding to the minimum value and OxFF corresponding to the maximum value are stored instead of the slope value obtained in the NUC table 261 when the pixel corresponds to the defective pixel. Ox00 and OxFF denoted by a red color indicate the values of the gain of normal pixels belonging to the normal range. The control unit of the system or the defective pixel processing unit 240 should know in advance that a pixel having a value of 0x00 and 0xFF is a defective pixel. This is an example of a minimum value and a maximum value, and may be set by various other methods, but is not limited thereto.

The defective pixel processing unit 240 corrects the numerical value by outputting values of adjacent pixels not corresponding to the minimum value and the maximum value when the slope value of each pixel of the column image data corresponds to a preset minimum value or maximum value. Referring to FIG. 6, the value 0x00 corresponding to the defective pixel is corrected by outputting the output value of the normal pixel positioned at the front or rear.

After the nonuniformity correction (NUC) table 261 is uploaded, the minimum pixel value and the maximum value information stored in the NUC table 261 may be transmitted to the bad pixel processing unit 240. Referring to FIG. 6, after the NUC table 261 is uploaded, Ox00 corresponding to the minimum value and OxFF corresponding to the maximum value may be transmitted to the defective pixel processing unit 240. FIG. By doing so, the defective pixel processing unit 240 knows in advance which value corresponds to the defective pixel.

Specifically, after the nonuniformity correction (NUC) table 261 is uploaded, the defective pixel processing unit 240 may not receive the minimum value and the maximum value information stored in advance in the NUC table 261, When the information indicating that there is no minimum value and no maximum value stored in the uploaded NUC table 261 is output without performing numerical correction on each pixel of the column image data without searching the uploaded NUC table 261. [ Since there is no positional recording of defective pixels in the NUC table 261, it is determined that all pixels are normalized and output without performing numerical correction for each pixel.

[Second Embodiment]

Hereinafter, another embodiment of the present invention will be described with reference to the accompanying drawings.

A detailed description of the parts overlapping with those of the first embodiment will be omitted.

The defective pixel detection apparatus 200 using the position memory of the thermal imaging camera according to the present invention includes an infrared sensor 210, a storage unit 260, a nonuniformity correction unit 230, and a defective pixel processing unit 240, And a post-processing unit 250. The post-

The infrared sensor 210 detects infrared rays emitted from the subject and outputs the infrared image data as thermal image data.

The storage unit 260 stores the non-uniformity correction (NUC) table 261 having the slope information using the output value of each pixel of the infrared sensor 210 while changing the temperature of the subject, A non-uniformity correction (NUC) table 261 having a predetermined minimum value or maximum value is stored instead of the corresponding slope.

The nonuniformity correction unit 230 refers to the uploaded NUC table 261 when the NUC table 261 stored in the storage unit 260 at the time of booting is uploaded and stores the NUC table 261 for each pixel of the thermal image data output from the infrared sensor 210 Correct the output value.

The defective pixel processing unit 240 searches the uploaded NUC table 261 for the slope value of each pixel of the column image data, corrects the value of the corresponding defective pixel if it corresponds to the predetermined minimum value or the maximum value, If the slope value of each pixel of the data does not correspond to the preset minimum value or maximum value, it outputs without numerical correction.

The storage unit 260 generates a non-uniformity correction (NUC) table 261 with a predetermined minimum value instead of the obtained slope when the output value of the defective pixel is smaller than the output value of the normal pixel. If the output value of the defective pixel is larger than the output value of the normal pixel A nonuniformity correction (NUC) table 261 can be generated with a predetermined maximum value instead of the obtained inclination.

Referring to FIG. 4, it can be seen that the output value of the defective pixel is shifted downward and upward. In the case where the output value is shifted downward, the output value is smaller than the output value of the normal pixel. Referring to FIG. 6, this case can be stored as a preset minimum value Ox00 instead of the obtained slope. The output value shifted upward is a case where the output value is larger than the output value of the normal pixel. Referring to FIG. 6, this case can be stored as a preset maximum value OxFF instead of the obtained slope.

The defective pixel processing unit 240 may correct the numerical value by outputting values of adjacent pixels not corresponding to the minimum value and the maximum value when the slope value of each pixel of the column image data corresponds to a preset minimum value or maximum value.

When the nonuniformity correction (NUC) table 261 stored in the boot-time storage unit 260 is uploaded, the control unit (not shown) transmits the minimum value and the maximum value information stored in advance in the NUC table 261 to the defective pixel processing unit 240 . By doing so, the defective pixel processing unit 240 knows in advance which value corresponds to the defective pixel.

Specifically, after the nonuniformity correction (NUC) table 261 is uploaded, the defective pixel processing unit 240 may not receive the minimum value and the maximum value information stored in advance in the NUC table 261, The NUC table 261 is not searched and outputted without performing numerical correction on each pixel of the column image data.

According to the present invention, since a separate additional memory for storing the position of the defective pixel DP is not required and a separate process for determining a defective pixel is not required in the mass production, the manufacturing cost and the production time can be saved have.

It will be apparent to those skilled in the art that many other modifications and applications are possible within the scope of the basic technical idea of the present invention.

200 ... Bad pixel detection device using position memory of thermal image camera
210 ... Infrared sensor
220 ... pre-
230 ... non-
240 ... bad pixel processing section
250 ... post processor
260 ... storage
261 ... Non-uniformity correction (NUC) table

Claims (11)

delete delete When generating a nonuniformity correction (NUC) table at the time of mass production, the positions of the defective pixels are recorded in advance in the nonuniformity correction table, and the nonuniformity correction table is uploaded at the time of booting the system so that the nonuniform correction unit and the defective pixel processing unit use the same A defective pixel detection method using position memory,
Obtaining an inclination using an output value of each pixel of the infrared sensor while changing a temperature of a subject;
Determining the pixel as a defective pixel if the slope is out of a predetermined range;
Generating a non-uniformity correction (NUC) table by storing a slope for each pixel in a storage unit to generate a non-uniformity correction (NUC) table, and storing a non-uniformity correction (NUC) table as a predetermined minimum or maximum value instead of the obtained gradient, The step of generating the nonuniformity correction (NUC) table may be performed by storing a predetermined minimum value instead of the obtained slope when the output value of the defective pixel is smaller than the output value of the normal pixel, and replacing the slope when the output value of the defective pixel is larger than the output value of the normal pixel Save as preset maximum value -;
Uploading a non-uniformity correction (NUC) table generated by the above at the boot time storage;
Transmitting the minimum value and maximum value information previously stored in the NUC table to the bad pixel processing unit;
Detecting infrared rays emitted from the subject by the infrared sensor and outputting the detected infrared rays as thermal image data;
Correcting an output value for each pixel of the column image data by referring to the uploaded NUC table; And
The defective pixel processing unit searches the uploaded NUC table for the slope value of each column of the column image data, corrects the value of the corresponding defective pixel if it corresponds to the preset minimum value or the maximum value, And outputting the corrected pixel value without performing the numerical correction if the pixel-by-pixel slope value does not correspond to the preset minimum value or maximum value.
The method of claim 3,
Wherein the defective pixel processing unit corrects the numerical value to an output value of an adjacent pixel not corresponding to a minimum value and a maximum value when the slope value of each pixel of the column image data corresponds to a predetermined minimum value or a maximum value, A defective pixel detection method using position memory of a video camera.
delete The method of claim 3,
After the nonuniformity correction (NUC) table is uploaded, the bad pixel processing unit transmits information indicating that the minimum value and the maximum value information stored in advance in the NUC table are not received, or that there is no minimum value and maximum value stored in advance in the NUC table Wherein the NUC table is output without performing a numerical correction on each pixel of the thermal image data without searching the uploaded NUC table.
When generating a nonuniformity correction (NUC) table at the time of mass production, the positions of the defective pixels are recorded in advance in the nonuniformity correction table, and the nonuniformity correction table is uploaded at the time of booting the system so that the nonuniform correction unit and the defective pixel processing unit use the same A defective pixel detecting apparatus using position memory,
An infrared sensor for detecting infrared rays emitted from a subject and outputting the infrared rays as thermal image data;
A non-uniformity correction (NUC) table having slope information using an output value of each pixel of the infrared sensor while changing a temperature of a subject is stored. In case of a defective pixel whose slope is out of a predetermined range, The storage unit for storing the nonuniformity correction (NUC) table having the maximum value generates a nonuniformity correction (NUC) table with a predetermined minimum value instead of the obtained slope when the output value of the bad pixel is smaller than the output value of the normal pixel, Generating a non-uniformity correction (NUC) table with a predetermined maximum value instead of the obtained gradient when the output value of the normal pixel is larger than the output value of the normal pixel;
A non-uniform correction unit for correcting an output value of each pixel of the thermal image data output from the infrared sensor by referring to the uploaded NUC table when the NUC table stored in the storage unit is uploaded at boot time; And
A slope value for each pixel of the thermal image data is retrieved from the uploaded NUC table, and if the slope value corresponds to a preset minimum value or a maximum value, a value of the corresponding defective pixel is corrected and output, And a defective pixel processing unit for outputting, without numerical correction, a predetermined minimum value or a maximum value,
Wherein when a nonuniformity correction (NUC) table stored in the storage unit is uploaded at boot time, the control unit transfers the minimum value and the maximum value information stored in advance in the NUC table to the bad pixel processing unit. Defective pixel detection device.
delete The method of claim 7,
Wherein the defective pixel processing unit corrects the numerical value to an output value of an adjacent pixel not corresponding to a minimum value and a maximum value when the slope value of each pixel of the column image data corresponds to a predetermined minimum value or a maximum value, A defective pixel detection apparatus using position memory of a video camera.
delete The method of claim 7,
After the nonuniformity correction (NUC) table is uploaded, the bad pixel processing unit transmits information indicating that the minimum value and the maximum value information stored in advance in the NUC table are not received, or that there is no minimum value and maximum value stored in advance in the NUC table And outputs the read NUC table without performing numerical correction on each pixel of the column image data without searching the uploaded NUC table.

Images