Infrared image bias output real-time correction method, storage medium and device

Abstract

The present application relates to a kind of infrared image bias output real-time correction method, storage medium and device, steps are as follows: the gain correction coefficient and bias correction coefficient of infrared image are obtained based on calibration method;Calculate the correction coefficient λ0 of current frame scene image and current bias coefficient;Calculate the space-time filtering value λ1 of correction coefficient;Calculate the local information amount I inf of image after non-uniform correction based on λ1;The correction coefficient of bias coefficient is updated again based on local information amount, and λ2 is obtained;The final corrected image is obtained based on λ2 and image non-uniformity correction.The present application can effectively improve the effect of non-uniform correction, and simultaneously prolong the stable time after single point correction;The algorithm complexity of this technique is low, and it is easy to realize on different platforms such as FPGA, DSP, ARM, PC etc.

Description

Technical Field

[0001] The present invention belongs to the field of infrared image processing technology, and relates to a real-time correction method for infrared image bias output, a storage medium and a device, and in particular to a real-time correction method for infrared image bias output based on spatiotemporal information. Background Art

[0002] Non-uniformity correction is one of the key technologies in infrared image processing. Currently, there are methods such as two-point calibration correction, multi-point calibration correction, polynomial correction, scene correction algorithm, or shutter correction combined. However, calibration-based correction algorithms will gradually degrade image quality with temperature drift; scene-based correction algorithms are highly dependent on the scene. For example:

[0003] CN109084899A discloses a method and device for correcting the non-uniformity of the bulk output of an infrared focal plane detector. This method predicts the bulk output values ​​at three consecutive temperature points at the same interval, which to a certain extent prolongs the image attenuation rate during continuous operation. However, when the linear range of the three consecutive temperature points is exceeded, the image quality will still be significantly attenuated.

[0004] CN109889694A discloses a parallel optimization system and method for SoC based on nonlinear correction of scene infrared images. This method is a real-time non-uniform correction system and method based on the scene. This method does not require shutter correction, but this type of method utilizes filtering and processing in the time domain, which is prone to ghosting when the scene changes rapidly. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to overcome the above-mentioned shortcomings and provide a real-time correction method, storage medium and device for infrared image bias output based on spatiotemporal information. On the basis of the traditional calibration correction algorithm, based on the combination of calibration and local optimization, the stabilization time after single-point correction is extended, which not only ensures the long-term stability of the image, but also prevents the occurrence of ghosting.

[0006] The technical solution adopted in the present invention is as follows:

[0007] A real-time correction method for infrared image bias output based on spatiotemporal information includes the following steps:

[0008] Step 1: Obtain the gain correction coefficient K and bias correction coefficient B of the infrared image based on the calibration method, where the gain correction coefficient can be obtained by two-point calibration, multi-point calibration, or a combination of multi-point calibration and curve fitting; the bias correction coefficient can be obtained by shutter correction, background correction, and defocus correction.

[0009] Step 2: Calculate the correction coefficient based on the current frame scene image and the current bias coefficient , calculation formula:

[0010] ;

[0011] in is the scene image, is the local window of the original image, is the local window mean of the original image, is the bias correction coefficient local window, is the local window mean of the bias correction coefficient, and the final The coefficients are the same size as the original image.

[0012] Further, the window size of 3×3~15×15 achieves better results. At the same time, for detection of some obvious horizontal and vertical stripes, one-dimensional signals can be used for correction, that is, 3×1~15×1 or 1×3~1×15.

[0013] Step 3: Calculate the time-space filter value of the correction coefficient , the calculation formula is as follows:

[0014] ;

[0015] in is the correction coefficient finally obtained by non-uniform correction of the previous frame image, is the time domain filtering parameter, The larger the value, the stronger the time domain filtering. The value range is between 0 and 1, and generally the best effect is achieved between 0.1 and 0.2. To represent filtering in the spatial domain, a low-pass filter such as a Gaussian kernel can be used. The size of the filter kernel is generally between 3×3 and 7×7 to achieve good results.

[0016] Step 4: Calculate based on Image after non-uniformity correction The local information , the calculation of local information can be obtained by using algorithms such as high-pass filter, guided filter layering to obtain detail layer, bilateral filter to obtain detail layer, gradient information, etc. The calculation formula is as follows:

[0017] ;

[0018] Step 5: Update the correction coefficient of the bias coefficient again based on the local information to obtain , which is calculated as follows:

[0019] ;

[0020] Further, The performance of the detector can also be measured by facing the imaging device to a uniform scene and continuously collecting The threshold is obtained by multiplying the images. The calculation formula is as follows:

[0021] ;

[0022] Step 6: Based on Perform image non-uniformity correction to obtain the final corrected image. The calculation formula is as follows:

[0023] ;

[0024] Update at the same time .

[0025] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which is executed by an FPGA, DSP, ARM or other processor to implement the steps of the infrared image bias output real-time correction method based on spatiotemporal domain information of the present invention.

[0026] The present invention also provides a real-time correction device for infrared image bias output, which includes processors on different platforms such as FPGA, DSP, ARM, PC, etc. and the computer-readable storage medium of the present invention.

[0027] The present invention also provides a simple black pot suppression device for infrared images, comprising a processor and the computer-readable storage medium of the present invention.

[0028] The beneficial effects of the present invention are:

[0029] 1) Effectively improve the effect of non-uniformity correction and extend the stabilization time after single-point correction;

[0030] 2) It not only ensures the long-term stability of the image, but also prevents the occurrence of ghosting;

[0031] 3) The algorithm complexity of this technology is low, and it is easy to implement on different platforms such as FPGA, DSP, ARM, and PC. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 The present invention is a flow chart of a method for real-time correction of infrared image bias output based on spatiotemporal information.

[0033] Figure 2 This is a comparison diagram of the images before and after correction using the method of the present invention. It can be seen from the figure that the image effect after correction is cleaner than the image before correction, the details are more obvious, and the image quality is significantly improved. DETAILED DESCRIPTION

[0034] The present invention is described in further detail below with reference to the embodiments.

[0035] Those skilled in the art will understand that the following examples are intended to illustrate the present invention only and should not be construed as limiting the scope of the present invention. Where specific techniques or conditions are not specified in the examples, the techniques or conditions described in the literature in the art or in the product specifications were used. Materials or equipment used without manufacturer identification are commercially available conventional products.

[0036] To illustrate the specific implementation process, take the 640×512 resolution infrared camera as an example:

[0037] Step 1: Obtain the gain correction coefficient K[640, 512] and the bias correction coefficient B[640, 512] of the infrared image based on the calibration method.

[0038] The further gain correction coefficient adopts the two-point calibration method, and the calculation formula is as follows:

[0039] ;

[0040] in is the high temperature image response image, is the low temperature image response image, is the mean value of low temperature image.

[0041] The bias correction coefficient is obtained by shutter correction acquisition.

[0042] Step 2: Calculate the correction coefficient based on the current frame scene image and the current bias coefficient , the calculation formula is:

[0043]

[0044] in is the scene image, is the local window of the original image, is the local window mean of the original image, is the bias correction coefficient local window, is the local window mean of the bias correction coefficient, and the final The coefficients are the same size as the original image.

[0045] The size of the further window is taken as a column-entry 1D vector with size 5×1.

[0046] Step 3: Calculate the time-space filter value of the correction coefficient , the calculation formula is as follows:

[0047] ;

[0048] The time domain filtering parameters , A 5×5 Gaussian kernel is used, and the Gaussian kernel is as follows:

[0049] ;

[0050] Furthermore, in a specific infrared imaging system, the selection of time domain parameters should be adjusted according to the rate of change of temperature drift, and the filter kernel can use low-pass filter kernels of different scales and algorithms.

[0051] Step 4: Calculate based on Image after non-uniformity correction The local information ,The amount of local information can be calculated using algorithms such as ,highpass filter, guided filtering layering to obtain detail layer, bilateral filtering to obtain detail layer, ,gradient information, and so on.

[0052] Further, The local information of is guided by filtering, and the calculation formula is as follows:

[0053] ;

[0054] ;

[0055] Based on gradient information, the calculation formula is as follows:

[0056] ;

[0057] ;

[0058] The gradient kernel The scale can be determined according to the original image characteristics.

[0059] Further, The calculation formula is as follows:

[0060]

[0061] Step 5: Update the correction coefficient of the bias coefficient again based on the local information to obtain , which is calculated as follows:

[0062] ;

[0063] in It is the threshold used to distinguish scene information from image non-uniform information, and generally achieves good results between 5 and 20.

[0064] Further, The performance of the detector can also be measured by facing the imaging device to a uniform scene and continuously collecting The threshold is obtained by multiplying the images. The calculation formula is as follows:

[0065] ;

[0066] In particular, when resources and time are sufficient, the more frames collected, the closer the actual noise threshold is.

[0067] Step 6: Based on Perform image non-uniformity correction to obtain the final corrected image. The calculation formula is as follows:

[0068] ;

[0069] Update at the same time .

[0070] In particular, for an imaging system, the first frame of the image It needs to be initialized to 1 to ensure fast convergence of the algorithm.

[0071] A method for real-time correction of infrared image bias output based on spatiotemporal domain information. The computer program product includes a non-transitory readable storage medium and a computer program. The computer program is tangibly stored on the non-transitory readable storage medium. The computer program is executed by a processor in an FPGA, DSP, ARM or other computer to perform the steps of the real-time correction technology for infrared image bias output based on spatiotemporal domain information.

[0072] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the foregoing embodiments. The foregoing embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A real-time correction method for infrared image bias output, characterized in that: The steps include: Step 1: Obtain the gain correction coefficient K and bias correction coefficient B of the infrared image based on the calibration method; Step 2: Calculate the correction coefficient based on the current frame scene image and the current bias coefficient ; Step 3: Calculate the time-space filter value of the correction coefficient : ； in: The correction coefficient finally obtained by non-uniform correction of the previous frame image; is the time domain filtering parameter; Represents filtering in the spatial domain; Step 4: Calculate based on Image after non-uniformity correction The local information ,in: ,in, is the scene image; Local information Obtained using high-pass filtering, guided filtering, bilateral filtering or gradient information method; Step 5: Update the correction coefficient of the bias coefficient again based on the local information to obtain : ； in: It is the threshold used to distinguish scene information from image non-uniform information; Step 6: Based on Perform image non-uniformity correction to obtain the final corrected image I out : ； Update at the same time .

2. The real-time correction method according to claim 1, wherein: The gain correction coefficient K is obtained by two-point calibration, multi-point calibration or a combination of multi-point calibration and curve fitting; the bias correction coefficient B is obtained by shutter correction, background correction or defocus correction.

3. The real-time correction method according to claim 2, characterized in that : Correction factor described in step 2 The calculation formula is as follows: ； in: is the local window of the original image, is the local window mean of the original image, is the bias correction coefficient local window, is the local window mean of the bias correction coefficient.

4. The real-time correction method according to claim 1, wherein: As described in step 3 The value range is between 0 and 1.

5. The real-time correction method according to any one of claims 1 to 4, characterized in that: described The performance of the detector is measured by continuously collecting The threshold is obtained by multiplying the images. The calculation formula is as follows: ； in, Indicates the width of the image, Indicates the height of the image, Description The The first Rank pixels, express The first sub-image of any Rank pixels.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the steps of the infrared image bias output real-time correction method according to any one of claims 1 to 5.

7. A simple infrared image black pot suppression device, characterized in that: The method comprises a processor and a computer-readable storage medium as claimed in claim 6.

Images