Thermal imaging device and thermal imaging method

Abstract

A thermal imaging device includes a thermal detector device (100) that includes an array (102) of thermally sensitive pixels and a signal processing circuit (104) coupled to the detector device (100). The circuit (104) supports a background identifier (110) and a pixel classifier (112), the background identifier (110) including a general intensity identifier (114) and an expected background intensity calculator (116). The background identifier (110) receives pixel measurement data captured by the detector device (100) with respect to the pixels of the array (102), and the general intensity identifier (114) identifies a maximum number of substantially identical pixel intensity values from the pixel measurement data. The expected background intensity calculator (116) uses the maximum number of substantially identical pixel intensity values to generate a model of an expected background intensity level. The pixel classifier (112) uses the model to determine whether intensity measurements by the pixels (118) of the array (102) correspond to a background or an object in an image.

Description

Technical Field

[0001] This invention relates to thermal imaging apparatuses, such as thermal imaging apparatuses comprising thermally sensing pixel arrays that respectively receive and measure infrared radiation. The invention also relates to thermal imaging methods that belong to the type that, for example, use thermally sensing pixel arrays to receive and measure infrared radiation. Background Technology

[0002] In the field of infrared detection equipment, detector devices comprising thermal sensing pixel arrays are known. When using such a detector device, it is known to determine a so-called digital image threshold, which is an ideal value used to distinguish the background and one or more thermal objects in the field of view measured by the detector device. Due to the nature of the thermal sensing element array (typically a 24x32 pixel array), there are not enough measurement data points available to determine the digital image threshold.

[0003] Many known techniques for determining the digitization threshold are based on histogram calculations and analysis. In this regard, data compartments are typically used, such as 8, 12, or 16 data compartments. The highest and lowest measured temperatures are identified to determine the range of measured temperatures in the thermal image. The range of compartments is then determined based on the measured temperature range and the number of data compartments used. Once the compartment ranges have been determined, the thermal image is typically scanned, and for each pixel, the image is reconstructed by replacing each measured temperature with the identifier of the compartment where the temperature of that pixel decreases. After such preprocessing, various techniques can be employed to determine the digitization threshold.

[0004] In one example, a histogram is generated based on compartment measurement data, and a so-called balanced histogram technique is applied to the data. The balanced histogram technique finds a balance point in the accumulated data of the histogram after excluding certain measurement levels. Alternatively, a so-called maximum distance valley technique can be used, where the highest or peak temperature compartment is determined based on the histogram about the foreground, and the highest or peak temperature is determined based on the histogram about the background. A line is then projected between the two peaks, and a vertical line with the maximum distance to the histogram compartment value is calculated. It is assumed that the count of compartments intersecting the vertical line corresponds to a digitization threshold. This technique works well for low compartment content information in the foreground. The development of this technique (called Otsu's method) involves iterating over possible thresholds and calculating a pixel-level spread metric on both sides of the threshold, i.e., whether these pixels are foreground or background pixels. The goal of this technique is to minimize the spread of foreground and background pixel intensity values. When the minimum spread is found, the digitization threshold is considered to have been found.

[0005] However, the techniques described above cannot always produce reliable results. Summary of the Invention

[0006] According to a first aspect of the invention, a thermal imaging apparatus is provided, comprising: a thermal detector device including a thermally sensing pixel array; a signal processing circuit operatively coupled to the thermal detector device and including a background identifier and a pixel classifier, the background identifier including a general intensity identifier and an expected background intensity calculator; the background identifier being configured to receive pixel measurement data of an image captured by the thermal detector device with respect to pixels of the thermally sensing pixel array during use, and the general intensity identifier being configured to identify a maximum number of substantially identical pixel intensity values ​​from the received pixel measurement data; the expected background intensity calculator being configured to use the maximum number of substantially identical pixel intensity values ​​to generate a model of expected background intensity levels that can be referenced by pixels; and the pixel classifier being configured to use the model to determine whether an intensity measurement performed per pixel of the thermally sensing pixel array corresponds to the background of the image or an object in the image.

[0007] The background recognizer may further include a pixel threshold calculator; the maximum number of substantially identical pixel intensity values ​​may have a range of measured pixel intensities associated with them; the expected pixel intensity calculator may be configured to compute a gradient across the range of measured pixel intensities, the model including the computed gradient; the expected pixel intensity calculator may be configured to use the computed gradient to generate multiple expected pixel background intensity values ​​for the thermally sensitive pixel array; the pixel threshold calculator may be configured to compare the multiple expected pixel intensity values ​​with pixel measurement data to determine a temperature threshold; and the pixel classifier may be configured to classify the intensity measurements as corresponding to either background or object by comparing the intensity measurements with the temperature threshold, based on the pixels of the thermally sensitive pixel array.

[0008] A universal intensity identifier can be configured to sort the received pixel measurement data in a certain order to provide sorted pixel intensity measurement data.

[0009] The order can be ascending or descending.

[0010] A universal intensity recognizer can be configured to identify the maximum number of substantially identical pixel intensity values ​​from sorted pixel intensity measurement data.

[0011] The universal intensity recognizer can be configured to calculate multiple intensity differences by calculating the differences between consecutive intensity measurements of sorted pixel intensity measurement data.

[0012] A universal intensity recognizer can be configured to identify the maximum number of substantially identical pixel intensity values ​​by identifying a range of substantially constant differences among multiple intensity differences, and identifying the maximum range of substantially constant differences, thereby identifying the maximum number of substantially identical pixel intensity values.

[0013] The gradient across the range of measured pixel intensity can be calculated with respect to several pixel intensity values ​​of the sorted pixel intensity measurement data that are identified as the maximum number of substantially the same pixel intensity values.

[0014] The background recognizer can be configured to round the received pixel measurement data before the general intensity recognizer identifies the maximum number of substantially the same pixel intensity values ​​from the pixel measurement data.

[0015] The general strength identifier can be configured to analyze multiple strength differences and identify the lowest non-zero strength difference among the multiple strength differences.

[0016] The received pixel measurement data can be stored as a first vector; multiple intensity differences can be stored as a second vector; multiple expected pixel background intensity values ​​can be stored as a third vector; and the index of each of the first and third vectors can be mapped to the index of the thermal sensing pixel array.

[0017] The pixel classifier can be configured to identify a first boundary pixel index with respect to a maximum number of substantially identical pixel intensity values ​​of sorted pixel intensity measurement data, and can be configured to select a first expected pixel intensity value of a plurality of expected pixel intensity values ​​and a first pixel measurement data having respective pixel indices that increase in distance from the first boundary pixel index; the thermal sensing pixel array can generate a plurality of intensity measurement values ​​including the intensity measurement values; and the pixel classifier can also be configured to compare each of the plurality of intensity measurement values ​​with a temperature threshold.

[0018] The increment of the index of the first expected pixel intensity value and the index of the first pixel measurement data of multiple expected pixel intensity values ​​can be in a first direction away from the first boundary pixel index.

[0019] The pixel classifier can be configured to identify a second boundary pixel index with respect to a maximum number of substantially identical pixel intensity values ​​of sorted pixel intensity measurement data, and can be configured to select a second expected pixel intensity value of multiple expected pixel intensity values ​​and a second pixel measurement data having a pixel index of pixel measurement data that increases in distance from the second boundary pixel index so as to decrease in value; and the pixel classifier can also be configured to compare each of the multiple intensity measurements with another temperature threshold.

[0020] The increment of the index of the second expected pixel intensity value and the index of the second pixel measurement data of multiple expected pixel intensity values ​​can be in a second direction away from the second boundary pixel index.

[0021] One temperature threshold corresponds to the recognition of pixels in a thermal sensing pixel array that measures a temperature greater than the background temperature, while another temperature threshold corresponds to the recognition of pixels in a thermal sensing pixel array that measures a temperature less than the background temperature.

[0022] The temperature threshold may include adjustments made through noise margin.

[0023] Noise margin can be several standard deviations of the calculated noise level.

[0024] The signal processing circuitry may further include a filter configured to filter the sorted pixel intensity measurement data. The filter may be configured to perform a sliding window averaging.

[0025] According to a second aspect of the invention, a thermal imaging method is provided, the thermal imaging method comprising: a thermally sensing pixel array of a thermal detector device capturing an image from received infrared radiation and generating pixel measurement data therefrom; a signal processing circuit receiving the pixel measurement data and identifying a maximum number of substantially identical pixel intensity values ​​from the received pixel measurement data; the signal processing circuit using the maximum number of substantially identical pixel intensity values ​​to generate a model of a expected background intensity level that can be referenced by the pixels; and the signal processing circuit using the model to determine whether an intensity measurement performed on a pixel-by-pixel basis corresponds to the background of the image or an object in the image.

[0026] According to a third aspect of the invention, a thermal imaging method is provided, the thermal imaging method comprising: capturing an image from received infrared radiation and generating pixel measurement data therefrom; receiving the pixel measurement data and identifying a maximum number of substantially identical pixel intensity values ​​from the received pixel measurement data; using the maximum number of substantially identical pixel intensity values ​​to generate a model of a expected background intensity level that can be referenced by the pixels; and using the model to determine whether the intensity measurement performed on a pixel-by-pixel basis of a thermally sensing pixel array corresponds to the background of the image or to an object in the image.

[0027] Therefore, an apparatus and method can be provided to detect pixels corresponding to the background of a scene in an image captured by a thermal detector device (e.g., a thermal imaging camera) with improved accuracy. The apparatus and method also improve the consistency of accurate detection of background pixels. Furthermore, the apparatus and method provide such results even when the sensing array comprises a small number of pixels (e.g., a 24x32 pixel array). Attached Figure Description

[0028] Referring to the accompanying drawings, at least one embodiment of the invention will now be described by way of example only, in which:

[0029] Figure 1This is a schematic diagram of a thermal detector device constituting an embodiment of the present invention;

[0030] Figure 2 It is by Figure 1 A screenshot of an image captured by a thermal detector device;

[0031] Figure 3 This is a flowchart of a thermal imaging method constituting another embodiment of the present invention;

[0032] Figure 4 This is a schematic diagram of the first vector of sorted pixel intensity values;

[0033] Figure 5 yes Figure 3 The method uses a sorted graph of measured pixel intensity values ​​and pixel differences; and

[0034] Figure 6 Through Figure 3 The method involves processing and identifying the pixel values ​​of the background and object pixels, resulting in a screenshot. Detailed Implementation

[0035] Throughout the following description, the same reference numerals will be used to identify the same parts.

[0036] refer to Figure 1 The thermal detector device 100 of the thermal imaging apparatus includes a thermal sensing pixel array 102 operatively coupled to a signal processing circuit 104. In this example, the thermal sensing pixel array 102 is arranged as a rectangular matrix of M columns and N rows of pixels. An example of a suitable thermal sensing pixel array is the MLX90640 far-infrared thermal sensor array available from Melexis Corporation, Nevada. The signal processing circuit 104 includes an image capture module 106 operatively coupled to the thermal sensing pixel array 102. Since a detailed description of the structure and / or operation of the image capture module 106 is not essential for understanding this or other embodiments, it will not be described in further detail herein for the sake of brevity and simplicity.

[0037] In this example, the signal processing circuitry 104 also includes data storage, such as memory 108, including random access memory (RAM), a background recognizer unit 110, and a pixel classifier unit 112. The image capture module 106 is operatively coupled to memory 108. Similarly, background recognizer unit 110 and pixel classifier unit 112 are operatively coupled to memory 108. Background recognizer unit 110 supports and therefore includes a general intensity recognizer unit 114, a pixel threshold calculator unit 115, and an expected background intensity calculator unit 116.

[0038] In operation ( Figure 2 The thermal detector device 100 operates in an environment where a thermal image needs to be captured. The thermal detector device 100 is oriented towards the scene in that environment and captures a thermal image 300. Figure 3 (Step 200). Infrared electromagnetic radiation emitted or reflected by objects in the scene is received by the thermal sensing pixel array 102 within the field of view of the thermal detector device 100. In this example, the captured thermal image 300 contains thermal representations of the object 302 (e.g., a person) and the background 304.

[0039] Infrared electromagnetic radiation received by the thermal sensing pixel array 102 is converted from the optical domain to the electrical domain. The image capture module 106 receives an electrical signal from the thermal sensing pixel array 102, representing raw intensity measurement data of the infrared electromagnetic radiation received by each pixel 118 of the thermal sensing pixel array 102. Using this received raw intensity data, the image capture module 106 generates temperature measurement data for each individual pixel 118 of the thermal sensing pixel array 102. The generated temperature measurement data is stored in the memory 108. As mentioned above, the thermal sensing pixel array 102 is arranged as an M x N matrix of pixels 118, and in this example, the temperature measurement data is stored in the memory 108 by row and column indexes of the thermal sensing pixel array 102, e.g., T. m,n , where T is the temperature of the pixel in the m-th column and n-th row.

[0040] Then, the general intensity identifier unit 114 accesses the temperature measurement data stored in memory 108 and sorts the temperature measurement data in ascending order (step 202). Of course, in other examples, the temperature measurement data can be sorted in descending order. In any case, in this example, the sorted temperature measurement data is then stored in the first vector 400.

[0041] In some embodiments, the temperature values ​​stored in the first vector 400 may be rounded (step 204) to the desired accuracy or within the capability of the signal processing circuit 104, for example, to 0.1 Kelvin. In another optional stage, the rounded (or unrounded if no rounding is performed) sorted measured temperature values ​​stored in the first vector 400 may be filtered (step 206) by a filter (not shown) of the signal processing circuit 104 using any suitable technique (e.g., sliding window averaging or any other suitable low-pass or averaging filter).

[0042] refer to Figure 5 The sorted temperature measurement data stored in the first vector 400, when plotted to improve visualization to help explain the example, results in an increase in the measured temperature in the first graph 500.

[0043] Using the sorted measured temperature data stored in memory 108 by the first vector 400, the general intensity identifier unit 114 then calculates (step 208) the temperature difference between consecutive adjacent indices of temperature stored in the first vector 400, i.e., T. n+1 –T n Where T is the stored temperature value and n is the index of the first vector 400. The calculated temperature difference values ​​of consecutive adjacent indices of the first vector 400 are stored (step 210) as a second vector (not shown). When plotted to aid visualization, the second temperature difference graph 502 includes measurements of temperature increments 504 between adjacent indices of the first vector 400. In this respect, since the intensity values ​​of the background 304 of the captured image 300 are generally fairly uniform in terms of temperature variation, especially after filtering, the second temperature difference graph 502 includes a number of substantially identical temperature difference values. Therefore, the purpose of the general intensity recognizer is to identify the maximum number of substantially identical pixel intensity values ​​from the temperature measurement data stored in the memory 108. In one example, this is achieved by the general intensity recognizer unit 114 identifying (step 212) the minimum non-zero temperature difference value 506 stored in the second vector. Typically, the second vector will include more than one occurrence of the minimum non-zero temperature difference value 506, and therefore include more than one pixel range with substantially identical intensity values. Once all minimum non-zero temperature differences 506 have been identified, the expected background intensity calculator unit 116 calculates (step 214) the index distances between each consecutive adjacent pairs of minimum non-zero temperature differences 506 with respect to the first vector 400. Once the index distances have been calculated, the longest distance 508 between the minimum non-zero temperature differences 506 is selected (step 216) to identify the largest range of pixels with substantially the same (unchanging) intensity values ​​associated with them, as mentioned above.

[0044] The longest distance 508 is defined by two minimum non-zero temperature differences 506 occurring at the lower index value i1 510 and the higher index value i2 512. The lower index value i1 510 has a corresponding first measured temperature value t1, which can be referenced and obtained from the sorted temperature measurement data stored by the first vector 400. Similarly, the second index value i2 512 has a corresponding second measured temperature value t2, which can be referenced and obtained from the sorted temperature measurement data stored by the first vector 400. Although substantially the same, this maximum number of substantially identical pixel intensity values ​​includes the range of the measured pixel intensity associated with them. The expected background intensity calculator unit 116 then uses the upper and lower index values ​​i1, i2 and their corresponding measured temperature values ​​t1, t2 to calculate (step 218) the gradient across the range of the measured pixel intensity:

[0045] Then, the expected background intensity calculator unit 116 uses the calculated gradient grad temp To calculate the expected background pixel temperature relative to the range of indices corresponding to all pixel coordinates: T Ei =(i-i1)grad temp +t1 (2)

[0046] Where i is the estimated background temperature T being calculated. Ei The current index, i, corresponds to the index of the first vector 400, i1 is the minimum value of the lower index mentioned above, and t1 is the first measured temperature value at the lower index value i1 mentioned above. Gradient grad temp The calculation and the application of equation (2) above constitute the modeling of the expected background intensity level through pixel reference.

[0047] In this example, equation (2) above is used by iterating over the index value i to calculate (step 220) the expected background pixel temperature value T corresponding to each index value i of the first vector 400. Ei The calculated expected background pixel temperature value is stored as a third vector in memory 108 (step 220). Therefore, in this example, it should be understood that the index of one or more of the first, second, and third vectors corresponds to or maps to the pixel index of the thermal sensing pixel array 102.

[0048] Once the expected background pixel temperature value T has been calculated Ei (Step 220) The pixel threshold calculator unit 115 then iterates through the indices of the first and third vectors in such a manner as to determine a first threshold th related to a temperature greater than the temperature of the background 304. h And a second threshold th related to a temperature smaller than the background temperature of 304. l To determine the first threshold th h The threshold calculator 115 selects the index value i2 and incrementally compares (step 222) the measured temperature value T stored in the first vector. i The expected background pixel temperature value T corresponding to the index of the third vector Ei The process of comparing the stored values ​​of the corresponding indices of the first and third vectors continues until the index value h is identified, where the measured temperature T... h Exceeding the corresponding expected background pixel temperature value T Eh A pre-calculated noise margin, the details of which are described below, is then used. The temperature at which this noise margin is taken into account is then recorded as the first threshold th. hSimilarly, in order to determine the second threshold th l The threshold calculator 115 selects the lower index value i1 and compares the measured temperature value T stored in the first vector in descending order of the index (step 224). i The expected background pixel temperature value T corresponding to the index of the third vector. Ei The process of comparing the stored values ​​of the corresponding indices of the first and third vectors continues until another index value l is identified, where the measured temperature T... l Compared to the corresponding expected background pixel temperature value T El A low pre-calculated noise margin is used. Then, the temperature at which the noise margin is taken into account is recorded as the second threshold th. l .

[0049] Although, in the example above, the first and second thresholds th h ,th l The search begins with the upper index value i2 and the lower index value i1 respectively, but it should be understood that in other examples, the pixel threshold calculator 115 can select index i. m index i m This corresponds to the midpoint of the longest distance 508 defined by the lower index value i1510 and the upper index value i2512, and then indexes i from the midpoint. m Search in decreasing and / or increasing directions.

[0050] Subsequently, in this example, pixel classifier unit 112 accesses the temperature measurement data initially generated by image capture module 106 and stored in memory 108, and compares (step 226) the temperature measurement data for each pixel stored in memory 108 with a first calculated threshold th. h To analyze data about each pixel of the thermal sensing pixel array 102. In this regard, the pixel classifier unit 112 constructs a first mapping corresponding to the measured temperature data for the thermal sensing pixel array 102 stored in the memory 108. This is done by determining when the measured temperature exceeds a first calculated threshold. h The first mapping is generated by storing an indication that the corresponding pixel of the thermal sensing pixel array 102 is a so-called "hot" pixel in the first mapping of the thermal sensing pixel array 102. Once the threshold th calculated for the first time is completed... h The test, then, is conducted in a similar manner with respect to the second calculated threshold th. l Mapping generation is performed. In this regard, pixel classifier unit 112 revisits the temperature measurement data initially generated by image capture module 106 and stored in memory 108, and compares (step 228) the temperature measurement data for each pixel stored in memory 108 with the second calculated threshold th. hThe data for each pixel of the thermal sensing pixel array 102 is analyzed. This comparison process is reused by the pixel classifier unit 112 to construct a second map, which in this example directly corresponds to the measured temperature data for the thermal sensing pixel array 102 stored in memory 108. The second map determines when the measured temperature is less than a second calculated threshold th. l And in the second mapping of the thermal sensing pixel array 102, an indication that the corresponding pixel of the thermal sensing pixel array 102 is a so-called "cold" pixel is generated.

[0051] In the example above, the first and second thresholds th are respectively. h ,th l Two mappings are generated. However, depending on the intended application of the device, only one of the first and second mappings may be required. Therefore, it should be understood that although the calculation of both the first and second thresholds and the subsequent generation of the mappings have been described above, for some applications, only one of the two thresholds may be calculated, and thus only one corresponding mapping of the two mappings may be generated.

[0052] In this example, the first and second predetermined thresholds th l ,th h It is used in conjunction with noise margin estimation (e.g., average noise margin). The average noise margin can be adjusted by a predetermined amount. In this respect, the margin can be based on several standard deviations from the average noise margin, such as 6 sigma. The average noise margin can be based on a calculation of the overall noise of the device. Threshold th l ,th h Alternatively, the number of items can be increased or decreased arbitrarily for optimization purposes based on the obtained results.

[0053] Once the pixels of captured image 300 are classified, further processing of captured image 300 is possible, taking into account pixels classified as background 600 and pixels classified as object 602. The nature of further processing is not central to understanding the example illustrated herein, and therefore will not be described in further detail for the sake of clarity and brevity.

[0054] In the example above, those skilled in the art will understand that any suitable memory management technique can be employed to enable the classification pixel data to be referenced back to temperature measurement data for each pixel 118 of the thermal sensing pixel array 102 generated by the image capture module 106 and stored in the memory 108. By employing such a technique, it is possible to correlate the results of pixel classification with respect to vector indices with the pixels 118 of the thermal sensing pixel array 102.

[0055] Those skilled in the art should understand that the implementation described above is merely an example of various implementations conceivable within the scope of the appended claims. In fact, it should be understood that in the examples above, the first and second predetermined thresholds th... l ,th h Different, but in other examples, the first and second predetermined thresholds th l ,th h The same threshold can be used; that is, a single threshold can be employed. It should also be understood that although the examples described herein illustrate the use of thermopile sensors, any suitable type of temperature sensor, such as a calorimeter, can be used.

[0056] Terms such as “greater than”, “greater than”, and “less than” mentioned in this article should also be understood to take into account alternative terms such as “greater than or equal to” and / or “less than or equal to”.

[0057] It should be understood that, unless otherwise explicitly stated, references to “infrared” herein are intended to refer to electromagnetic energy (including electromagnetic radiation in the near-infrared, mid-infrared, and far-infrared regions) with wavelengths ranging from 700 nm to 1 mm (such as 760 nm to 1 mm or 700 nm to 1500 nm).

Claims

1. A thermal imaging device, comprising: A thermal detector device, the thermal detector device comprising a thermal sensing pixel array; A signal processing circuit, operatively coupled to the thermal detector device, includes a background recognizer and a pixel classifier, the background recognizer including a general intensity recognizer and an expected background intensity calculator; The background recognizer is configured to receive pixel measurement data of an image captured by the thermal detector device with respect to the pixels of the thermal sensing pixel array during use, and the universal intensity recognizer is configured to identify a maximum number of substantially identical pixel intensity values ​​from the received pixel measurement data, wherein the maximum number of substantially identical pixel intensity values ​​includes a range of measured pixel intensities associated with them. The expected background intensity calculator is configured to use the maximum number of substantially the same pixel intensity values ​​to generate a model of the expected background intensity level that can pass through a pixel reference; as well as The pixel classifier is configured to use the model to determine whether an intensity measurement performed on the pixels of the thermally sensing pixel array corresponds to the background of the image or an object in the image.

2. The apparatus of claim 1, wherein, The background recognizer further includes a pixel threshold calculator; The expected background intensity calculator is configured to calculate a gradient across a range of the measured pixel intensities, and the model includes the calculated gradient. The expected background intensity calculator is configured to use the calculated gradient to generate multiple expected pixel background intensity values ​​for the thermally sensing pixel array; The pixel threshold calculator is configured to compare the plurality of expected pixel background intensity values ​​with the pixel measurement data in order to determine a temperature threshold; as well as The pixel classifier is configured to classify the intensity measurement as corresponding to the background or the object by comparing the intensity measurement with the temperature threshold, according to the pixels of the thermally sensing pixel array.

3. The apparatus of claim 2, wherein the universal intensity identifier is configured to sort the received pixel measurement data in a certain order to provide sorted pixel intensity measurement data.

4. The apparatus of claim 3, wherein the universal intensity identifier is configured to identify the maximum number of substantially identical pixel intensity values ​​from the sorted pixel intensity measurement data.

5. The apparatus of claim 3, wherein the universal intensity identifier is configured to calculate a plurality of intensity differences by calculating the differences between consecutive intensity measurements of the sorted pixel intensity measurement data.

6. The apparatus of claim 5, wherein the universal intensity recognizer is configured to recognize the maximum number of substantially identical pixel intensity values ​​by: recognizing a range of substantially constant differences among the plurality of intensity differences, and recognizing the maximum range of the substantially constant differences, thereby recognizing the maximum number of substantially identical pixel intensity values.

7. The apparatus of claim 6, wherein the universal strength identifier is configured to analyze the plurality of strength differences and identify the lowest non-zero strength difference among the plurality of strength differences.

8. The apparatus of claim 5, wherein, The received pixel measurement data is stored as a first vector; The multiple intensity differences are stored as a second vector; The multiple expected pixel background intensity values ​​are stored as a third vector; as well as The index of each of the first vector and the third vector is mapped to the index of the thermal sensing pixel array.

9. The apparatus of claim 4, wherein, The pixel classifier is configured to identify a first boundary pixel index with respect to the maximum number of substantially identical pixel intensity values ​​of the sorted pixel intensity measurement data, and is configured to select a first expected pixel background intensity value of the plurality of expected pixel background intensity values ​​and a first pixel measurement data of the pixel measurement data having respective pixel indices that increase in distance from the first boundary pixel index; The thermal sensing pixel array generates multiple intensity measurement values, including the intensity measurement values; and The pixel classifier is also configured to compare each of the plurality of intensity measurements with the temperature threshold.

10. The apparatus of claim 9, wherein, The pixel classifier is configured to identify a second boundary pixel index with respect to the maximum number of substantially identical pixel intensity values ​​of the sorted pixel intensity measurement data, and is configured to select a second expected pixel background intensity value of the plurality of expected pixel background intensity values ​​and a second pixel measurement data having respective pixel indices that increase with a distance from the second boundary pixel index so that the values ​​decrease; and The pixel classifier is also configured to compare each of the plurality of intensity measurements with another temperature threshold.

11. The apparatus of claim 2, wherein the temperature threshold includes an adjustment made by means of noise margin.

12. The apparatus of claim 11, wherein the noise margin is a plurality of standard deviations of the calculated noise level.

13. The apparatus of claim 3, wherein the signal processing circuitry further includes a filter configured to filter the sorted pixel intensity measurement data.

14. The apparatus of claim 3, wherein the filter is configured to perform sliding window averaging.

15. The apparatus of claim 1, wherein the universal intensity identifier is configured to sort the received pixel measurement data in a certain order to provide sorted pixel intensity measurement data.

16. The apparatus of claim 15, wherein the universal intensity identifier is configured to identify the maximum number of substantially identical pixel intensity values ​​from the sorted pixel intensity measurement data.

17. The apparatus of claim 15, wherein the universal intensity identifier is configured to calculate a plurality of intensity differences by calculating differences between consecutive intensity measurements of the sorted pixel intensity measurement data.

18. The apparatus of claim 17, wherein the universal intensity recognizer is configured to recognize the maximum number of substantially identical pixel intensity values ​​by: recognizing a range of substantially constant differences among the plurality of intensity differences, and recognizing the maximum range of the substantially constant differences, thereby recognizing the maximum number of substantially identical pixel intensity values.

19. The apparatus of claim 15, wherein the signal processing circuitry further includes a filter configured to filter the sorted pixel intensity measurement data.

20. The apparatus of claim 15, wherein the filter is configured to perform a sliding window averaging.

21. A thermal imaging method, comprising: The thermal detector device's thermal sensing pixel array captures images from received infrared radiation and generates pixel measurement data from them; The signal processing circuit receives the pixel measurement data and identifies a maximum number of substantially identical pixel intensity values ​​from the received pixel measurement data, wherein the maximum number of substantially identical pixel intensity values ​​includes a range of measured pixel intensities associated with them; The signal processing circuit uses the maximum number of substantially the same pixel intensity values ​​to generate a model that can pass the expected background intensity level of the pixel reference; and The signal processing circuit uses the model to determine whether the intensity measurement performed on the pixels of the thermal sensing pixel array corresponds to the background of the image or an object in the image.

Images