Description:TECHNICAL FIELD
[0001] The present disclosure relates, in general, to imaging systems, and more specifically, relates to a system and method for the automatic selection of non-uniformity correction (NUC) table for thermal imagers.

BACKGROUND
[0002] Thermal imagers, extensively used in surveillance, industrial inspections, and medical diagnostics, heavily rely on accurate sensor calibration to ensure uniformity across the entire imaging system. However, due to technological constraints, such as limited data acquisition capabilities and inconsistent gain in column amplifiers, non-uniformities persist in thermal images, which can affect system performance over longer distances.
[0003] A few examples of such systems are recited in U.S. Patent no. 9208542 B2, which discusses the reduction of fixed pattern noise in thermal imagers by receiving an image frame arranged in multiples of rows and columns. Non-uniformity correction is achieved by processing image data to obtain column correction terms which are obtained by the relative relation between the pixels in the corresponding column and its neighbourhood within a selected distance from the corresponding pixels.
[0004] Another example is recited in U.S. Patent no. 9508124 B2, which describes the method of shutter-less correction of infrared imagers. This method involves the generation of a correction image array and the detection of edges in the correction array. This method receives the image from the camera and the corrected image based on the correction from the previous frame. This method checks for the detection of edges in the corrected image when the imager is moving and thus updates the correction table to result in fixed pattern noise correction.
[0005] Another example is recited in U.S. Patent no. 8760509 B2, which describes the removal of fixed-pattern noise artefacts. This method discusses the correction of the non-uniformity present in the infrared imager using shutter and shutter-less methods. This method discusses the offset compensation method for non-uniformity correction. But during power on of thermal imager, the temperature inside the imager gets heated up, therefore, becoming less thermally stable. Hence the shutter operation needs to be performed at a higher rate resulting in more actuation thus leading to the end user’s inconvenience. The advantages and disadvantages of shutter-based non-uniformity correction have also been brought out in the description.
[0006] Another example is recited in U.S. Patent no. 9167179 B2, which describes a method where the calibration of a micro-bolometer-based focal plane array will be carried out by the processing unit with the help of an external computer. This method discusses the acquisition of sensor data and the process involved in calibration with the acquisition of temperature frame data and the computation involved in removing the non-uniformity. Another example is recited in U.S. Patent no. 9723227 B2, which describes the method for correction of fixed pattern noise with the blurred image frames. Their main idea was to figure out the noise by finding the differences between the blurred image and desired scene data. Here, the image frames are intentionally obtained by the defocusing the lens. These are employed to find the correction terms in the column and row respectively.
[0007] Yet another example is recited in U.S. Patent no. 9292909 B2, which describes the non-uniformity correction with respect to a vehicle in the scene. Sensor data is acquired in association with the motion of the vehicle. These sensor signals are processed to determine the motion of the imaging device with respect to the scene. This parameter is used to perform image correction. It also discusses the correction with the deployment of the shutter and also lists the downside of using the shutter.
[0008] Therefore, it is desired to overcome the drawbacks, shortcomings, and limitations associated with existing solutions, and develop a system that provides automatic selection of NUC tables to ensure that the thermal imaging system operates at its optimal performance level.

OBJECTS OF THE PRESENT DISCLOSURE
[0009] An object of the present disclosure relates, in general, to an imaging system, and more specifically, relates to a system and method for the automatic selection of non-uniformity correction (NUC) table for thermal imagers.
[0010] Another object of the present disclosure is to provide a system that provides automatic selection of NUC tables to ensure that the thermal imaging system operates at its optimal performance level. By selecting the most suitable NUC table for each scenario, potential non-uniformities and distortions in the thermal images are corrected, resulting in enhanced image quality and accuracy.
[0011] Another object of the present disclosure is to provide a system that significantly reduces the need for manual intervention in the NUC table selection process. By leveraging automated algorithms and computational techniques within the processing unit, the system can autonomously determine and apply the appropriate NUC table for a given scenario. This reduces human error and streamlines the operation of the thermal imager.
[0012] Another object of the present disclosure is to provide the automation of the NUC table selection process saving valuable time and resources. Without the need for manual calibration and selection, system operators can focus on other tasks, leading to increased productivity.
[0013] Another object of the present disclosure is to provide a system that eliminates the potential costs associated with human error in selecting the NUC tables, improving overall efficiency and cost-effectiveness.
[0014] Another object of the present disclosure is to provide a system that considers the specific imaging conditions and requirements of each scenario. By dynamically selecting the NUC table based on these factors, the system can adapt to different environmental conditions, such as changes in temperature or ambient lighting, ensuring optimal performance and accurate thermal imaging results.
[0015] Yet another object of the present disclosure is to provide a system that enhances the user experience with thermal imagers. The resulting improved image quality and accuracy provide users with clearer, more reliable thermal information, leading to better decision-making and analysis in various applications, including surveillance, industrial inspections, and medical diagnostics.

SUMMARY
[0016] The present disclosure relates in general, to the imaging system, and more specifically, relates to a system and method for the automatic selection of Non-Uniformity Correction (NUC) table for thermal imagers. The main objective of the present disclosure is to overcome the drawback, limitations, and shortcomings of the existing system and solution, by providing a system and method for automatic selection of a Non-uniformity correction (NUC) table for thermal imagers. The basic system comprises of imager unit, processing unit and display unit. Acquired image sensor data is passed onto the processing unit. Automatic selection of the NUC table for any given scenario occurs at the processing unit and the corresponding processor unit output is coupled onto the display. This methodology helps in minimising the human intervention effectively leading to optimal system performance.
[0017] The method for automatically selecting the Non-Uniformity Correction (NUC) table for thermal imagers, addresses the challenges associated with the fabrication process of thermal imagers, particularly the Focal Plane Array. The method aims to mitigate non-uniformities in the imager caused by limitations in pixel data acquisition and column amplifier gain variations. The proposed automated method leverages advancements in technology to mitigate non-uniformity issues. By intelligently selecting the most suitable NUC table, the method minimizes human intervention and maximizes system performance. This automated approach optimizes image quality by reducing non-uniformities and enhances the imager's ability to capture targets even at longer distances, surpassing the limitations posed by non-uniformity in the read-out process.
[0018] The present disclosure provides a processing unit that is coupled to the image sensing unit, the processing unit configured to acquire the raw data from the image sensing unit at defined temperatures. The processing unit can compute the gain value and offset values at different temperatures from the acquired raw data, compute different NUC tables with different integration times in different environmental conditions, and select the appropriate region of interest for the selection of the NUC table. The region of interest is chosen as the portion surrounding the central M*N region to eliminate the target which is positioned in the center of the image. Further, processing unit 104 can compute the pixel average of the selected region of interest, compare the computed pixel average with low and high thresholds and proceed with the corresponding NUC table and the integration time for generating the final output image if the computed average falls within the specified threshold.
[0019] The processing unit changes the integration time of the image sensing unit to shift the histogram towards the center of the dynamic range of the thermal imager when the spatial average falls outside the threshold range. The processing unit performs a comparison check to figure out whether the spatial average falls within the specified threshold range. The processing unit, upon the condition not being met, performs a comparison between the pixel average and the low threshold, and in the case where the pixel average is found to be lower than the low threshold, the processing unit proceeds to increase the integration time of the thermal imager. The processing unit proceeds to decrease the integration time of the imager when the pixel average is greater than or equal to the low threshold range. The pixel average of the region lies within the acceptable range after changing the integration time is verified. When the pixel average is still outside the acceptable range, continue the process for other integration times. The integration time of the image sensing unit is adjusted to ensure that the current scene dynamic range falls within the desired window of the full dynamic range of the image sensing unit. If the condition is satisfied, the respective NUC table is derived for mapping.
[0020] Various objects, features, aspects, and advantages of the inventive subject matter will become more apparent from the following detailed description of preferred embodiments, along with the accompanying drawing figures in which like numerals represent like components.

BRIEF DESCRIPTION OF THE DRAWINGS
[0021] The following drawings form part of the present specification and are included to further illustrate aspects of the present disclosure. The disclosure may be better understood by reference to the drawings in combination with the detailed description of the specific embodiments presented herein.
[0022] FIG.1 illustrates the general block diagram of an imaging system in accordance with an embodiment of the present disclosure.
[0023] FIG. 2 illustrates the block diagram of the processing operation of the system in accordance with an embodiment of the present disclosure.
[0024] FIG. 3 illustrates the flow chart for the automatic selection of the NUC table in accordance with an embodiment of the present disclosure.
[0025] FIG. 4 illustrates the flow chart of a method for the automatic selection of the NUC table in accordance with an embodiment of the present disclosure.

DETAILED DESCRIPTION
[0026] The following is a detailed description of embodiments of the disclosure depicted in the accompanying drawings. The embodiments are in such detail as to clearly communicate the disclosure. If the specification states a component or feature “may”, “can”, “could”, or “might” be included or have a characteristic, that particular component or feature is not required to be included or have the characteristic.
[0027] As used in the description herein and throughout the claims that follow, the meaning of “a,” “an,” and “the” includes plural reference unless the context clearly dictates otherwise. Also, as used in the description herein, the meaning of “in” includes “in” and “on” unless the context clearly dictates otherwise.
[0028] The present disclosure relates, in general, to imaging system, and more specifically, relates to a system and method for the automatic selection of the Non-Uniformity Correction (NUC) table for thermal imagers. The present invention introduces a method for the automatic selection of Non-Uniformity Correction (NUC) tables in thermal imagers. The system consists of an image sensing unit, processing unit, and display unit, and aims to minimize human intervention while maximizing system performance. The acquired image sensor data is transmitted to the processing unit for further analysis and correction. The automatic selection of the appropriate NUC table for a given scenario takes place within the processing unit. This selection process is designed to optimize the quality and accuracy of the thermal images produced.
[0029] By leveraging advanced algorithms and computational techniques, the processing unit identifies the most suitable NUC table based on the specific imaging conditions and requirements. This automatic selection eliminates the need for manual intervention, enhancing efficiency and reducing potential human errors in the NUC table selection process. The output from the processing unit, which incorporates the selected NUC table, is then coupled to the display unit for visualization. The display unit presents the final processed image to the user, ensuring optimal image quality and system performance.
[0030] The methodology introduces an innovative approach to automate the selection of NUC tables in thermal imagers. By minimizing human intervention and relying on advanced algorithms, the system achieves optimal performance while delivering accurate and high-quality thermal images. This advancement holds great potential for various applications where thermal imagers are utilized, including surveillance, industrial inspections, and medical diagnostics. The present disclosure can be described in enabling detail in the following examples, which may represent more than one embodiment of the present disclosure.
[0031] The term “raw signal/data” in the disclosure refers to the digital representation of analog electrical signals obtained from analog to digital converter from each of the sensing element. It represents the brightness level measured by the detector.
[0032] The term “noise” in the disclosure refers to the fixed pattern noise generated in the signal imagery because of the column amplifiers of the Read-Out Integrated Circuit.
[0033] The present disclosure highlights the challenges associated with non-uniformity in thermal imagers, which lead to fixed pattern noise in infrared (IR) images. The non-uniformity arises from physical variations among the pixels in the Focal Plane Array (FPA) and can be influenced by temperature changes within the FPA or the ambient environment. To achieve optimal system performance, it is essential to address the non-uniform response of the sensing elements, characterized by varying gain and offset values. In an ideal scenario, the sensing elements in the FPA should exhibit a linear response to impinging photons, ensuring uniformity across the image array. However, in practice, each sensing element displays its own unique gain and offset characteristics, with the offset values drifting over time. To achieve a linear response and improve system performance, a correction method is employed to adjust the gain and offset values of each pixel.
[0034] The objective of the correction process is to model the required signal output of each sensing element, accounting for both gain and offset corrections. By calibrating the gain and offset values for each pixel, a linear response can be achieved, thereby reducing non-uniformity and enhancing overall system performance. The desired outcome is to approach unit gain and zero offset, representing an ideal pixel response. Implementing non-uniformity correction techniques addresses the inherent non-uniformity in thermal imagers, effectively mitigating fixed pattern noise. By achieving a linear response across all sensing elements, this correction method significantly improves image quality, accuracy, and the overall performance of thermal imaging systems.
[0035] The advantages achieved by the system of the present disclosure can be clear from the embodiments provided herein. The method offers a significant advancement in thermal imaging technology, ensuring improved performance, increased accuracy, and enhanced imaging capabilities, while reducing reliance on manual intervention. The description of terms and features related to the present disclosure shall be clear from the embodiments that are illustrated and described; however, the invention is not limited to these embodiments only. Numerous modifications, changes, variations, substitutions, and equivalents of the embodiments are possible within the scope of the present disclosure. Additionally, the invention can include other embodiments that are within the scope of the claims but are not described in detail with respect to the following description.
[0036] FIG.1 illustrates the general block diagram of the imaging system in accordance with an embodiment of the present disclosure.
[0037] Referring to FIG. 1, imaging system 100 (also referred to as a system 100, herein) enables the acquisition, processing, and visualization of images. The system 100 can include an image sensing unit 102, a processing unit 104 and a display unit 106. The image sensing unit 102 captures the initial input image, the processing unit 104 applies various image processing algorithms to enhance and manipulate the image, and the display unit 106 presents the final processed image for viewing and analysis.
[0038] The image sensing unit 102 is responsible for capturing the input image from the scene. It typically consists of a sensor or an array of sensors, such as a Charge-Coupled Device (CCD) or a Complementary Metal-Oxide-Semiconductor (CMOS) sensor, which convert incoming optical signals into electrical signals. These electrical signals represent the intensity or color information of the captured image.
[0039] The processing unit 104 is coupled to the image sensing unit 102 which acquires the raw data from the image sensing unit 102 at defined temperatures. The processing unit 104 can compute the gain value and offset values at different temperatures from the acquired raw data, compute different NUC tables with different integration times in different environmental conditions, and select the appropriate region of interest for the selection of the NUC table. The region of interest is chosen as the portion surrounding the central M*N region to eliminate the target which is positioned in the center of the image. Further, processing unit 104 can compute the pixel average of the selected region of interest, compare the computed pixel average with low and high thresholds and proceed with the corresponding NUC table and the integration time for generating the final output image if the computed average falls within the specified threshold.
[0040] The processing unit 104 changes the integration time of the image sensing unit to shift the histogram towards the center of the dynamic range of the thermal imager, when the pixel average falls outside the threshold range. The processing unit 104, upon the condition not being met, performs a comparison between the pixel average and the low threshold, and in the case where the pixel average is found to be lower than the low threshold, the processing unit proceeds to increase the integration time of the thermal imager. The processing unit 104 proceeds to decrease the integration time of the imager when the pixel average is greater than or equal to the low threshold range. The pixel average of the region lies within the acceptable range after changing the integration time is verified. When the pixel average is still outside the acceptable range, continue the process for other integration times. The integration time of the image sensing unit is adjusted to ensure that the current scene dynamic range falls within the desired window of the full dynamic range of the image sensing unit. If the condition is satisfied, the respective NUC table is derived for mapping.
[0041] The processing unit 104 is responsible for performing various operations on the digital signals obtained from the image sensing unit 102. It typically includes a microprocessor or a dedicated image processing chip. The processing unit 104 may perform tasks such as noise reduction, image enhancement, color correction, geometric transformations, object recognition, and other image processing algorithms to improve the quality and usability of the captured image. It may also handle other functions such as image compression or data encoding.
[0042] The display unit 106 receives the processed output from the processing unit 104 and presents the final image to the viewer. It can be a screen, monitor, or any visual output device capable of displaying images. The display unit 106 converts the digital signals into a visual representation that can be perceived by the human eye. It may include technologies such as Liquid Crystal Displays (LCD), Organic Light Emitting Diodes (OLED), or other display technologies. The quality and characteristics of the display unit greatly influence the visual experience and the ability to accurately interpret the displayed image.
[0043] Thus, the present invention overcomes the drawbacks, shortcomings, and limitations associated with existing solutions, and provides a system that offers a significant advancement in thermal imaging technology, ensuring improved performance, increased accuracy, and enhanced imaging capabilities, while reducing reliance on manual intervention. The system automatically selects NUC tables for thermal imaging, reducing manual intervention, saving time and resources, eliminating potential costs of human error, considering specific imaging conditions, and enhancing the user experience.
[0044] FIG. 2 illustrates the block diagram of the processing operation of the system in accordance with an embodiment of the present disclosure.
[0045] The system can include acquisition unit 202 and pre-processor unit 206. The acquisition unit 202 is responsible for capturing the input raw data 204 from the scene. It may consist of sensors or detectors that convert physical signals such as light or radiation into electrical signals. The acquired analog data is then passed through an Analog-to-Digital Converter (A/D converter) to convert it into digital format, which can be further processed by the system. The raw data 204 represents the digital information obtained from the acquisition unit 202. It contains the original, unprocessed data captured from the scene. This raw data serves as the input for subsequent processing stages in the imaging system.
[0046] The pre-processor unit 206 receives the raw data and performs initial pre-processing tasks to enhance the quality and usability of the data. This unit typically includes algorithms for non-uniformity correction, which aims to reduce or eliminate the fixed pattern noise caused by variations among the pixels in the imaging system. It may also handle tasks such as bad pixel replacement, where defective pixels are identified and replaced with suitable values. The pre-processor unit prepares the data for further processing and analysis.
[0047] The processing unit 104 takes the pre-processed output from the pre-processor unit and performs more advanced processing operations on the data. This unit applies algorithms and techniques tailored to enhance specific aspects of the scene of interest. For example, it may employ detail enhancement algorithms to improve the visibility of fine details or features in the image. The processing unit can utilize various image processing techniques, such as filtering, noise reduction, contrast enhancement, and image sharpening, to enhance the overall quality and visual characteristics of the image.
[0048] The display unit 106 receives the processed data from the processing unit and encodes it into a suitable format for display. It may involve converting the digital data into signals compatible with the specific display device or encoding it into a desired image format. The display unit is responsible for presenting the processed image to the user on a display device, such as a monitor, screen, or other visual output medium. It ensures that the processed image is accurately rendered and visually perceivable to the user.
[0049] FIG. 3 illustrates the block diagram of the flow chart for automatic selection of NUC table in accordance with an embodiment of the present disclosure.
[0050] Referring to FIG. 3, the process 300 at block 302 involves acquiring sensor signals i.e., raw data from the imaging sensor and performing calculations to determine the optimal gain and offset parameters. At block 304, for a given integration time, computation of gain and offset values are obtained by projecting the imaging system to a uniform target of different temperatures for a desired temperature range of operation. Gain value can be obtained by the computation of the difference between the average value of the complete frame for a given temperature to the average value for another temperature over the difference between the corresponding pixels in first frame and second frame for their corresponding temperatures.
[0051] Computation of offset value can be obtained by the difference between the average value of the first frame and individual pixels in the second frame to the average value of the second frame and the individual pixels in the first frame over the difference between its corresponding pixels in first frame and second frame. The final resultant corrected image pixel can be a combination of these computed parameters. Different sets of tables were computed with different integration times for system operation in different environmental conditions.
[0052] The methodology basically deals with the automatic selection of correction tables for any particular atmospheric condition. This is achieved by changing the integration time of the imaging sensor. Basically, at block 306, the region of interest surrounding the centre portion of M*N pixels is chosen. The spatial average of the above-mentioned region for any given table is computed. At block 308, the computed spatial average of the given table is compared with the low and high thresholds. At block 310, if it lies within the range of the mentioned thresholds, then the system is left unchanged with the existing NUC table and integration time leading to the final output image. If the value is outside this range mentioned, then the integration time of the imaging sensor is changed to make its histogram shift in such a way that it lies almost in the centre of the dynamic range of the imager. At block 312, upon failing the condition, the pixel average is compared with the low threshold as mentioned above. At block 314, if the compared pixel average is less than the low threshold, then the integration time of the imager is increased else the integration time of the imager is decreased at block 316. Upon changing the integration time of the imaging sensor, the pixel average of the above region is again compared to check whether it lies within the acceptable range. If it does not, then the above-mentioned process is continued for other integration times. If the pixel average of the region lies within the low and high thresholds, then the integration time which was used to calibrate the particular NUC table is selected for the best optimal system performance. Human intervention is minimised by this procedure.
[0053] The proposed methodology offers an automated approach for selecting correction tables in imaging systems. By dynamically adjusting the integration time based on pixel averages and predefined thresholds, the system can optimize performance for varying atmospheric conditions, reducing the reliance on manual intervention.
[0054] FIG. 4 illustrates the flow chart of a method for the automatic selection of NUC table in accordance with an embodiment of the present disclosure.
[0055] The method 400 includes at block 402, the image sensing unit can capture the raw data from the scene at defined temperatures. At block 404, the processing unit can receive from the image sensing unit, the raw data at defined temperatures. At block 406, the processing unit can compute from the acquired raw data, the gain value and offset values at different temperatures. At block 408, the processing unit can compute different NUC tables with different integration times in different environmental conditions, At block 410, the appropriate region of interest is selected for selection of the NUC table. At block 412, the pixel average of the selected region of interest is computed. At block 414, the computed pixel average is compared with low and high thresholds. At block 416, proceed with the corresponding NUC table and the integration time for generating the final output image if the computed average falls within the specified threshold.
[0056] It will be apparent to those skilled in the art that the system 100 of the disclosure may be provided using some or all of the mentioned features and components without departing from the scope of the present disclosure. While various embodiments of the present disclosure have been illustrated and described herein, it will be clear that the disclosure is not limited to these embodiments only. Numerous modifications, changes, variations, substitutions, and equivalents will be apparent to those skilled in the art, without departing from the spirit and scope of the disclosure, as described in the claims.

ADVANTAGES OF THE PRESENT INVENTION
[0057] The present invention provides a system that provides an automatic selection of NUC tables to ensure that the thermal imaging system operates at its optimal performance level.
[0058] The present invention provides a system that significantly reduces the need for manual intervention in the NUC table selection process.
[0059] The present invention provides the automation of the NUC table selection process saving valuable time and resources.
[0060] The present invention provides a system that eliminates the potential costs associated with human error in selecting the NUC tables, improving overall efficiency and cost-effectiveness.
[0061] The present invention provides a system that considers the specific imaging conditions and requirements of each scenario.
[0062] The present invention provides a system that enhances the user experience with thermal imagers.
, Claims:
1. A system (100) for automatic selection of non-uniformity correction (NUC) tables in thermal imagers, the system comprising:
an image sensing unit (102) captures the raw data from the scene at defined temperatures; and
a processing unit (104) coupled to the image sensing unit, the processing unit configured to:
receive, from the image sensing unit, the raw data at defined temperatures;
compute, from the acquired raw data, the gain value and offset values at different temperatures;
compute different NUC tables with different integration times in different environmental conditions;
select the appropriate region of interest for the selection of the NUC table;
compute pixel average of the selected region of interest;
compare the computed pixel average with low and high thresholds; and
proceed with the corresponding NUC table and the integration time for generating the final output image if the computed average falls within the specified threshold.
2. The system as claimed in claim 1, wherein the processing unit (104) changes the integration time of the image sensing unit to shift the histogram towards the center of the dynamic range of the thermal imager, when the pixel average falls outside the threshold range.
3. The system as claimed in claim 1, wherein the processing unit (104), upon the condition not being met, performs a comparison between the pixel average and the low threshold, and in the case where the pixel average is found to be lower than the low threshold, the processing unit proceeds to increase the integration time of the thermal imager.
4. The system as claimed in claim 1, wherein the processing unit (104) proceeds to decrease the integration time of the thermal imager when the pixel average is greater than or equal to the low threshold range.
5. The system as claimed in claim 1, wherein the pixel average of the region lies within the acceptable range after changing the integration time is verified
6. The system as claimed in claim 1, wherein when the pixel average is still outside the acceptable range, continue the process for other integration times.
7. The system as claimed in claim 1, wherein the integration time of the image sensing unit is adjusted to ensure that the current scene dynamic range falls within the desired window of the full dynamic range of the image sensing unit.
8. The system as claimed in claim 7, wherein if the condition is satisfied, the respective NUC table is derived for mapping.
9. The system as claimed in claim 1, wherein the region of interest is chosen as the portion surrounding the central M*N region to eliminate the target which is positioned in the center of the image.
10. A method (400) for automatic selection of non-uniformity correction (NUC) tables in thermal imagers, the method comprising:
capturing (402), by an image sensing unit (102), the raw data from the scene at defined temperatures;
receiving (404), at a processing unit, from the image sensing unit, the raw data at defined temperatures;
computing (406), at the processing unit, from the acquired raw data, the gain value and offset values at different temperatures;
computing (408), at the processing unit, different NUC tables with different integration times in different environmental conditions;
selecting (410) the appropriate region of interest for selection of the NUC table;
computing (412) pixel average of the selected region of interest;
comparing (414) the computed pixel average with low and high thresholds; and
proceeding (416) with the corresponding NUC table and the integration time for generating the final output image if the computed average falls within the specified threshold.