CN111464755A - Signal-to-noise ratio real-time processing method and system for short-exposure image sequence of satellite-borne camera - Google Patents
Signal-to-noise ratio real-time processing method and system for short-exposure image sequence of satellite-borne camera Download PDFInfo
- Publication number
- CN111464755A CN111464755A CN202010311683.8A CN202010311683A CN111464755A CN 111464755 A CN111464755 A CN 111464755A CN 202010311683 A CN202010311683 A CN 202010311683A CN 111464755 A CN111464755 A CN 111464755A
- Authority
- CN
- China
- Prior art keywords
- data
- satellite
- short
- exposure
- memory
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/70—Circuitry for compensating brightness variation in the scene
- H04N23/73—Circuitry for compensating brightness variation in the scene by influencing the exposure time
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/80—Camera processing pipelines; Components thereof
- H04N23/81—Camera processing pipelines; Components thereof for suppressing or minimising disturbance in the image signal generation
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a method and a system for processing the signal-to-noise ratio of a short-exposure image sequence of a satellite-borne camera in real time, which relate to the technical field of image processing and comprise the following steps: collecting multi-frame continuous short exposure images; acquiring a plurality of satellite attitude data corresponding to a plurality of frames of continuous short-exposure images within shooting time; carrying out interpolation operation on the satellite attitude data to obtain satellite attitude data corresponding to multiple continuous short-exposure images one by one; carrying out attitude operation on the satellite attitude data subjected to interpolation operation to obtain correction data of multi-frame continuous short-exposure images which are in one-to-one correspondence with the attitude data subjected to interpolation operation; carrying out pixel address mapping calculation on pixels on the multi-frame continuous short-exposure images which correspond to the correction data one by one, and reading pixel address data to obtain enhanced image data; applying the enhanced image data; and clearing the data in the memory. According to the invention, the enhanced image data is obtained through the satellite-borne FPGA, so that the image contrast can be improved, and random noise can be eliminated.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to a method and a system for processing the signal-to-noise ratio of a short-exposure image sequence of a satellite-borne camera in real time.
Background
With the rapid development of computer graphics, imaging has been widely used in military, medical, commercial, and everyday life. In a special application environment, a common visible light image is limited by an imaging system and the light incoming amount of the surrounding environment, so that the difference between image noise and information contained in the image is small, the signal-to-noise ratio of the imaged image is too low, the target is blurred, and even the target is covered by background noise, and therefore, in practical application, the image with low signal-to-noise ratio needs to be subjected to image enhancement processing.
At present, the signal-to-noise ratio performance of a satellite image is more dependent on the dynamic range of an imaging device, and the performance of the device becomes a bottleneck limiting the satellite observation capability. In order to realize the overall improvement of the satellite image quality and the index requirement that the signal-to-noise ratio of the water color observation is superior to 1000, it is necessary to analyze the noise generation link and the noise suppression method in the imaging process in more detail and clarify the related technical means of the improvement of the signal-to-noise ratio.
Existing image enhancement methods include histogram equalization, which first counts the probability of each gray value appearing in the histogram, second accumulates the normalized histogram, and finally calculates new pixel values. The gray scale value with higher gray scale probability density in the image is expanded to the gray scale value with lower gray scale probability density nearby by using the cumulative function to adjust the gray scale value so as to realize contrast enhancement, so that the gray scale probability distribution in the image is changed and homogenized. The histogram equalization method can effectively improve the image contrast and expand the dynamic range of the gray scale. However, since the algorithm does not select the processed data in the statistical probability distribution, the contrast of the background and noise may be increased and the contrast of the target signal may be decreased, so that the gray value of the transformed image is decreased, causing some details to disappear, and making the finally displayed image unclear.
Therefore, in order to solve the problem of low imaging SNR due to limited optical aperture, detector sensitivity and one-time exposure time of the monitor, a method for superimposing short exposures of images is urgently needed to improve image contrast and eliminate random noise.
Disclosure of Invention
In view of the above, the invention provides a method and a system for processing a signal-to-noise ratio of a short-exposure image sequence of a satellite-borne camera in real time, wherein enhanced image data is acquired through a satellite-borne FPGA, so that image contrast can be improved, and random noise can be eliminated.
The application provides a signal-to-noise ratio real-time processing method for a short-exposure image sequence of a satellite-borne camera, which comprises the following steps:
acquiring a plurality of frames of continuous short-exposure images through a satellite-borne FPGA, and storing the plurality of frames of continuous short-exposure images into a first memory;
the satellite-borne FPGA is used for controlling an attitude sensor to acquire a plurality of satellite attitude data corresponding to a plurality of frames of continuous short-exposure images within shooting time, and the satellite attitude data are stored in a second memory;
carrying out interpolation operation on the satellite attitude data to obtain satellite attitude data corresponding to multiple frames of continuous short-exposure images one by one, and storing the satellite attitude data into a third memory;
carrying out attitude operation on the satellite attitude data subjected to interpolation operation to obtain corrected data of multi-frame continuous short-exposure images which correspond to the attitude data subjected to interpolation operation one by one, and storing the corrected data into a fourth memory;
carrying out pixel address mapping calculation on pixels on the multi-frame continuous short-exposure images which correspond to the correction data one by one, and reading pixel address data until all the pixel address data on the multi-frame continuous short-exposure images are read, so as to obtain enhanced image data;
transmitting the enhanced image data obtained by processing to any equipment for further processing or converting the enhanced image data into a video signal and outputting the video signal to imaging equipment;
and clearing the data in the first memory, the second memory, the third memory and the fourth memory.
Optionally, the operation method of the interpolation operation is: and inputting the horizontal and vertical coordinates of the attitude data and an interpolation enabling signal to the FPGA by adopting a secondary interpolation mode, and performing corresponding double-precision floating point number multiplication and division on the horizontal and vertical coordinates of the input attitude data and the interpolation enabling signal by adopting a floating point number operation IP core of the Sailing technology by the FPGA.
Optionally, the operation method of the gesture operation is: inputting a plurality of parameters to the FPGA, wherein the parameters comprise corresponding satellite-borne FPGA shooting angles, attitude data, Gaussian templates and enabling signals, and performing operation by adopting a double-precision floating point type operation method.
Optionally, inputting the corresponding parameters and outputting the corresponding results of the posture calculation are performed in a pipeline mode.
Optionally, the corrective data includes a translation matrix and a rotation matrix.
Optionally, pixel address mapping calculation is performed on pixels on multiple frames of continuous short-exposure images corresponding to the correction data one by one, and pixel address data is read until all pixel address data on the multiple frames of continuous short-exposure images are read, so as to obtain enhanced image data, specifically:
and sequentially carrying out pixel address mapping calculation on each line of pixels in the multi-frame continuous short-exposure images in one-to-one correspondence to the correction data, reading one line of pixel address data after completing one line of pixel address mapping calculation, carrying out superposition operation, and completing the calculation of the whole image until the pixel address data on the multi-frame continuous short-exposure images in one-to-one correspondence to the correction data are all completed, so that the enhanced image data is obtained.
Alternatively, the capacity of the pixel address is 2 times the number of pixels included in one row of pixels.
Optionally, before the enhanced image data obtained by processing is transmitted to other devices for further processing, or is converted into a video signal and output to an imaging device for display, the method includes:
and storing the enhanced image obtained by processing, wherein the image storage size is 10 bits.
Optionally, the pose data includes three-dimensional spatial information including pitch, pan and roll angles and temporal information.
The application also provides a signal-to-noise ratio real-time processing system based on the short-exposure image sequence of the satellite-borne camera, and the system comprises:
the acquisition module I is used for acquiring a plurality of frames of continuous short-exposure images through the satellite-borne FPGA and storing the plurality of frames of continuous short-exposure images into a first memory;
the acquisition module II is used for acquiring a plurality of satellite attitude data corresponding to the acquisition of a plurality of frames of continuous short-exposure images within shooting time through the satellite-borne FPGA control attitude sensor and storing the plurality of satellite attitude data into a second memory;
the first operation module is used for carrying out interpolation operation on the satellite attitude data to obtain satellite attitude data corresponding to multiple frames of continuous short-exposure images one by one and storing the satellite attitude data into a third memory;
the second operation module is used for carrying out attitude operation on the satellite attitude data subjected to interpolation operation to obtain corrected data of multi-frame continuous short-exposure images corresponding to the attitude data one by one, and storing the corrected data into a fourth memory;
the operation module III is used for carrying out pixel address mapping calculation on pixels on the multi-frame continuous short-exposure images which correspond to the correction data one by one, and reading pixel address data until all the pixel address data on the multi-frame continuous short-exposure images are read, so as to obtain enhanced image data;
the application module is used for transmitting the enhanced image data obtained by processing to any equipment for further processing or converting the enhanced image data into a video signal and outputting the video signal to the imaging equipment;
and the processing module is used for emptying the data in the first memory, the second memory, the third memory and the fourth memory.
Compared with the prior art, the method and the system for processing the signal-to-noise ratio of the short-exposure image sequence of the satellite-borne camera in real time at least realize the following beneficial effects:
1. in the satellite-borne camera short exposure image sequence signal-to-noise ratio real-time processing method and system, a plurality of frames of continuous short exposure images are collected through a satellite-borne FPGA, a plurality of satellite attitude data are collected through an attitude sensor controlled by the satellite-borne FPGA, interpolation operation is carried out on the satellite attitude data, the satellite attitude data corresponding to the plurality of frames of continuous short exposure images one by one are obtained, the change of the satellite attitude data caused by the shaking of shooting time can be eliminated, and the stability under an abnormal state is improved.
2. In the method and the system module for processing the signal-to-noise ratio of the short-exposure image sequence of the satellite-borne camera in real time, the signal-to-noise ratio of the image can be effectively improved through superposition of multiple continuous short-exposure images, and the signal-to-noise ratio improvement multiple is increased along with the increase of the superposed frame number, but due to the registration deviation of the image, the signal-to-noise ratio improvement multiple does not completely meet the requirement of the signal-to-noise ratio
The law of the multiplication is reduced to a certain extent.
3. According to the method and the system for processing the signal-to-noise ratio of the short-exposure image sequence of the satellite-borne camera in real time, the first storage, the second storage, the third storage and the fourth storage are called to store image data in the image processing process, and a pipeline type processing mode is used for the image processing process, so that the multi-step synchronous processing is realized, the image calculation speed can be greatly increased, and the hardware programming complexity and the memory space occupied by the system are reduced.
4. According to the method and the system for processing the signal-to-noise ratio of the short-exposure image sequence of the satellite-borne camera in real time, a hardware system based on an FPGA technology has great flexibility, and the difficulty of pcb design and clock design is reduced by using the FPGA for image processing.
Of course, it is not necessary for any product in which the present invention is practiced to achieve all of the above-described technical effects simultaneously.
Other features of the present invention and advantages thereof will become apparent from the following detailed description of exemplary embodiments thereof, which proceeds with reference to the accompanying drawings.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description, serve to explain the principles of the invention.
FIG. 1 shows a flow chart of a method for processing a signal-to-noise ratio of a short-exposure image sequence of a satellite-borne camera in real time according to an embodiment of the present application;
FIG. 2 is a flow chart of another method for processing the signal-to-noise ratio of a short-exposure image sequence of a satellite-borne camera in real time according to an embodiment of the present application;
FIG. 3 is a flowchart illustrating a signal-to-noise ratio real-time processing method for a short-exposure image sequence of a satellite-borne camera according to an embodiment of the present application;
fig. 4 shows a schematic structural diagram of a signal-to-noise ratio real-time processing system based on a short-exposure image sequence of a satellite-borne camera according to an embodiment of the present application.
Detailed Description
Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be noted that: the relative arrangement of the components and steps, the numerical expressions and numerical values set forth in these embodiments do not limit the scope of the present invention unless specifically stated otherwise.
The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application, or uses.
Techniques, methods, and apparatus known to those of ordinary skill in the relevant art may not be discussed in detail but are intended to be part of the specification where appropriate.
In all examples shown and discussed herein, any particular value should be construed as merely illustrative, and not limiting. Thus, other examples of the exemplary embodiments may have different values.
It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, further discussion thereof is not required in subsequent figures.
The following detailed description is to be read in connection with the drawings and the detailed description.
Example one
Fig. 1 is a flowchart of a method for processing a signal-to-noise ratio of a short-exposure image sequence of a space-borne camera in real time according to an embodiment of the present disclosure, and as shown in fig. 1, the method for processing a signal-to-noise ratio of a short-exposure image sequence of a space-borne camera in real time according to the present disclosure includes:
step S101, collecting multiple frames of continuous short-exposure images through a satellite-borne FPGA, and storing the multiple frames of continuous short-exposure images into a first storage;
step S102, controlling an attitude sensor to collect a plurality of satellite attitude data corresponding to the shooting time of a plurality of continuous short-exposure images through a satellite-borne FPGA, and storing the satellite attitude data into a second memory;
step S103, carrying out interpolation operation on the satellite attitude data to obtain satellite attitude data corresponding to multiple frames of continuous short-exposure images one by one, and storing the satellite attitude data into a third memory;
step S104, carrying out attitude operation on the satellite attitude data after the interpolation operation to obtain corrected data of multi-frame continuous short-exposure images which are in one-to-one correspondence with the attitude data after the interpolation operation, and storing the corrected data into a fourth memory;
step S105, pixel address mapping calculation is carried out on pixels on the multi-frame continuous short-exposure images which correspond to the correction data one by one, and pixel address data are read until all pixel address data on the multi-frame continuous short-exposure images are read, so that enhanced image data are obtained;
step S106, transmitting the enhanced image data obtained by processing to any equipment for further processing, or converting the enhanced image data into a video signal and outputting the video signal to imaging equipment;
and step S107, emptying the data in the first memory, the second memory, the third memory and the fourth memory.
Specifically, a plurality of frames of continuous short-exposure images are collected through a satellite-borne FPGA (Field-Programmable Gate Array), the satellite-borne FPGA controls a frame type CCD camera to store the plurality of frames of continuous short-exposure images into a first memory according to rows, a plurality of satellite attitude data are collected through a satellite-borne FPGA control attitude sensor while the satellite-borne FPGA shoots the plurality of frames of continuous short-exposure images and are stored into a second memory, interpolation operation is carried out on the plurality of satellite attitude data to obtain satellite attitude data corresponding to the plurality of frames of continuous short-exposure images one by one, attitude operation is carried out on the plurality of frames of continuous short-exposure images to obtain a translation matrix and a rotation matrix of pixels on each image, namely multi-frame continuous short-exposure image correction data corresponding to the satellite attitude data one by one, if the plurality of frames of continuous short-exposure images comprise N pieces of correction data, N groups of correction data are corresponded, and the correction data are stored into a fourth memory, and performing pixel address calculation on pixels on the multi-frame continuous short-exposure images corresponding to the corrected data to obtain enhanced image data, applying the enhanced image data, and emptying all memories for reuse after obtaining the enhanced image data.
It should be noted that the interpolation operation is performed by a newton interpolation method, and a function f (x) is used to make a suitable specific function based on the function values of a plurality of points in a certain interval, and the values of the specific function are taken as the approximate values of the function f (x) at the points and other points in the interval. The newton interpolation method has a successive advantage over the lagrange interpolation method in that the previous operation result can be used to reduce the operation amount when adding additional interpolation points.
According to the method for processing the signal-to-noise ratio of the short-exposure image sequence of the satellite-borne camera in real time, a hardware system based on an FPGA technology has great flexibility, and the difficulty of pcb design and clock design is reduced by using the FPGA for image processing; the satellite attitude data corresponding to the multi-frame continuous short exposure images one by one is obtained by acquiring multi-frame continuous short exposure images through the satellite-borne FPGA, controlling the attitude sensor to acquire a plurality of satellite attitude data through the satellite-borne FPGA and carrying out interpolation operation on the satellite attitude data, so that the change of the satellite attitude data caused by the shake of shooting time can be eliminated, and the stability in an abnormal state is improved; in addition, in the image processing process, the first memory, the second memory, the third memory and the fourth memory are called to store the image data, and the image processing process adopts a pipeline type processing mode, so that the multi-step synchronous processing is realized, the image calculation speed can be greatly improved, and the hardware programming complexity and the memory space occupied by the system are reduced.
Optionally, in step S103, the interpolation operation is performed by: and inputting the horizontal and vertical coordinates of the attitude data and an interpolation enabling signal to the FPGA by adopting a secondary interpolation mode, and performing corresponding double-precision floating point number multiplication and division on the horizontal and vertical coordinates of the input attitude data and the interpolation enabling signal by adopting a floating point number operation IP core of the Sailing technology by the FPGA.
Specifically, the interpolation operation is performed according to a data stream form, if 16 frames of images are superimposed, 16 times of interpolation operation is required, 16 satellite attitude data and corresponding parameters are sequentially input into the interpolation module according to a predetermined sequence, and after a time delay, the interpolation module continuously outputs 16 interpolated attitudes.
Optionally, in step S104, the attitude calculation method includes: inputting a plurality of parameters to the FPGA, wherein the parameters comprise corresponding satellite-borne FPGA shooting angles, attitude data, Gaussian templates and enabling signals, and performing operation by adopting a double-precision floating point type operation method.
Specifically, the objective of the attitude calculation is to obtain correction data, perform attitude calculation on all the interpolated satellite attitude data, and obtain a plurality of correction data, where each correction data corresponds to one image and one satellite attitude data.
Optionally, inputting the corresponding parameters and outputting the corresponding results of the posture calculation are performed in a pipeline mode.
Specifically, data input and output are performed in a pipeline mode, namely a series of parameters needing to be calculated can be continuously input, and corresponding image posture transformation data can be quickly obtained.
Optionally, the corrective data obtained in step S104 includes a translation matrix and a rotation matrix.
Specifically, in the attitude operation process, pixels in the images are translated and rotated, each image comprises a plurality of pixel points, and the position of each pixel point is translated and rotated to obtain a translation matrix and a rotation matrix.
Optionally, in step S105, pixel address mapping calculation is performed on pixels in the multiple frames of continuous short-exposure images corresponding to the correction data one to one, and pixel address data is read until all pixel address data in the multiple frames of continuous short-exposure images are read, so as to obtain enhanced image data, specifically:
and sequentially carrying out pixel row address mapping calculation on each row of pixels in the multi-frame continuous short-exposure images in one-to-one correspondence to the correction data, reading one row of pixel address data after completing one row of pixel address mapping calculation, carrying out superposition operation, and completing the whole image calculation until the pixel address data on the multi-frame continuous short-exposure images in one-to-one correspondence to the correction data are all completed, so as to obtain enhanced image data.
Specifically, correction data stored in the fourth memory are sequentially read, pixel address calculation is performed on pixels on multiple continuous short-exposure images corresponding to the correction data one by one, each image is sequentially calculated, after the pixel address calculation of one line corresponding to each image is completed, data in the pixel address of each line are read, then superposition operation is performed until the pixel addresses of all lines on each image are completely read, and then the pixel address calculation of the next image is performed until all image operation is completed.
It should be noted that, each line of pixel address data is read, and then the superposition operation is performed, which is performed in a data stream manner.
Alternatively, the capacity of the pixel address is 2 times the number of pixels included in one row of pixels.
Specifically, the capacity of the pixel address is set to be 2 times of the number of pixels in one row of pixels, so that the pixel address is not overflowed, the pixel address is not fully written, and when the pixel address data is larger than the number of pixels in one row of pixels, the pixel address data in one row is read, so that the pixel address is not read to be empty, and the reliability of the system is ensured.
Optionally, before step S106, that is, before the enhanced image data obtained by the processing is transmitted to another device for further processing, or is converted into a video signal and output to the imaging device for display, the method includes:
and storing the enhanced image obtained by processing, wherein the image storage size is 10 bits.
Specifically, after the application interface and the design of interactive operation are completed, the images shot by the camera and processed images need to be stored, the size of the image storage is 10 bits, the size of the display images in the system is 8 bits, and the program has the real-time demonstration capability.
Optionally, the attitude data acquired in step S102 includes three-dimensional space information and time information, where the three-dimensional space information includes a pitch angle, a pan angle, and a roll angle.
Specifically, each attitude data comprises three-dimensional space information and time information, the three-dimensional space information comprises pitching, translating and overturning angles, and the three-dimensional space information is calculated when attitude data interpolation is carried out, so that the three-dimensional space information corresponds to multiple frames of continuous short-exposure images one by one.
The present invention will be further described with reference to specific examples.
Example two
Fig. 2 is a flowchart of another method for processing a signal-to-noise ratio of a short-exposure image sequence of a satellite-borne camera in real time according to an embodiment of the present disclosure, and as shown in fig. 2, the method for processing a signal-to-noise ratio of a short-exposure image sequence of a satellite-borne camera in real time according to another embodiment of the present disclosure includes:
step S201, controlling the frame type camera to collect multiple frames of continuous short-exposure images through the satellite-borne FPGA, and storing the images into a memory DDR2 according to lines.
Step S202, a satellite-borne FPGA is used for controlling an attitude sensor to collect a plurality of satellite attitude data corresponding to the shooting time of a plurality of continuous short-exposure images, and the collected satellite attitude data are stored in a memory RAM1, wherein the satellite attitude data have three-dimensional spatial information, and the spatial information comprises pitching, translating and overturning angles.
Step S203, carrying out interpolation operation on the satellite attitude data by adopting a Newton interpolation method to obtain satellite attitude data corresponding to multiple continuous short-exposure images one by one, and storing the satellite attitude data into a memory RAM2, wherein the Newton interpolation method uses function values of a plurality of points in a certain interval of a function f (x) to make a proper specific function, the known values are taken at the points, and the values of the specific function are used as approximate values of a function f (x) at other points in the interval.
Specifically, the newton interpolation method has a successive advantage over the lagrange interpolation method in that the previous operation result can be utilized to reduce the operation amount when an additional interpolation point is added.
And step S204, attitude operation is carried out on the satellite attitude data after interpolation operation, the attitude operation is carried out according to the angle of translation and rotation of pixels in each image after interpolation, the correction data of multi-frame continuous short-exposure images in one-to-one correspondence is calculated according to the satellite attitude data stored in the third memory, if N images exist, N groups of data are calculated, and the calculated data are stored in the RAM 3.
And step S205, sequentially reading the corrected data in the RAM3, calculating according to pixel rows in each image, reading the data in the address after completing the calculation of one row of pixel addresses, superposing the data, performing the calculation of the second row of data after completing the calculation of one row of pixel addresses in the whole image, and performing the calculation of the next image until the calculation of all the images is completed.
And step S206, overlapping the calculated image with the image stored in the memory DDR2 to obtain enhanced image data.
And step S207, transmitting the enhanced image data obtained by processing to any equipment according to a protocol standard for processing continuously, or converting the enhanced image data into a video signal and outputting the video signal to imaging equipment.
Step S208, clearing all the data in the memory.
EXAMPLE III
Fig. 3 is a flowchart of a method for processing a signal-to-noise ratio of a short-exposure image sequence of a satellite-borne camera in real time according to an embodiment of the present disclosure, and as shown in fig. 3, the method for processing a signal-to-noise ratio of a short-exposure image sequence of a satellite-borne camera in real time according to another embodiment of the present disclosure includes:
s301, selecting a satellite-borne FPGA as a core of an SNR image processing system, controlling a frame type CCD camera to shoot a group of multi-frame continuous short-exposure images by the satellite-borne FPGA, transmitting the multi-frame continuous short-exposure images to the satellite-borne FPGA through T L K2711, performing ping-pong operation in the satellite-borne FPGA through two asynchronous FIFOs for data exchange, and storing the multi-frame continuous short-exposure images into a first memory in the ping-pong operation process.
It should be noted that Virtex V5L X110TFPGA from Xilinx is adopted as a core device of the SNR image processing system.
In the step, aiming at the problems of the frame type CCD camera, the problems of less image information due to small light input quantity and larger noise in the machine due to long exposure time are processed, the superposition processing of the continuous images in a short time is completed by using a short-exposure multi-frame superposition method, the image has higher signal-to-noise ratio compared with the images before processing or other processing methods, and the information of the image can be better maintained.
And S302, controlling an attitude sensor to acquire a plurality of satellite attitude data corresponding to the shooting time of the multi-frame continuous short-exposure image through the satellite-borne FPGA, and storing the satellite attitude data into a second memory.
Specifically, the satellite attitude data and the multiple frames of continuous short-exposure images are not completely matched, and a series of processing is required to acquire required data in order to acquire more accurate satellite attitude data.
The satellite attitude data includes point-by-Point Weighted Polynomial (PWPM), and the attitude and orbit at a known time can be used, and the time difference between the interpolation time and the set known time can be used as a metric to determine the weight, and the polynomial coefficients of the attitude and orbit at the time to be interpolated are calculated by the least square method, and the external orientation element at the time can be calculated by using the coefficients.
The weight of the interpolated time instant to each known time instant may be calculated as the inverse of the absolute value of the time difference or the inverse of the square of the absolute value of the time difference.
As exemplified below, when using 4 known times t on the track1、t2、t3、t4And when interpolating an external orientation element at a certain time t, since polynomial coefficients used for interpolation of the attitude and the orbit at the same time are the same, a calculation formula is given by taking a kappa (k) angle as an example, wherein the interpolation calculation formula is:
wherein v is an interpolation result,
and kappatRespectively, the kappa (k) angle of the unused time, t being a certain time.
4 error equations can be listed through kappa (k) angles of 4 known moments, and the weight p from the t moment to the known moment is calculated1、p2、p3、p4And forms a weight matrix P. And constructing a normal equation coefficient matrix N-1 and a normal equation free term vector U by a least square method based on the weight matrix P. For the convenience of derivation, N is-1And the contents of the U matrix are expressed in the form:
wherein, c1～c9Is a normal equation coefficient matrix N-1Term of (1), u1～u3Is a term in the normal equation free term vector U.
Calculating other coefficients according to the least square principle, and calculating kappa (k) value at the time t by using the calculated coefficients, wherein the calculation formula is as follows:
k(t)＝(c1u1+c2u2+c3u3)+(c4u1+c5u2+c6u3)t+(c7u1+c8u2+c9u3)t2wherein k (t) is the value of kappa (k) angle.
After obtaining the required satellite attitude information, a suitable model may be selected for geometric correction of the multiple frames of consecutive short-exposure images, the model used being described below.
The model takes a one-dimensional affine projection image as a base, and provides a correction mode of 'center projection-affine projection', namely, after a line center projection image is converted into a corresponding affine projection image by using a geometric relation during imaging, the spatial position is corrected by using a control point on the basis of the affine image. The correction test of the model for SPOT 1-level and 2-level stereo images proves that the correction can be completed only by 6 uniformly distributed ground control points, the plane precision can reach 6m and the elevation precision can reach 7.5m after the geometric correction.
And step S303, carrying out interpolation operation on the satellite attitude data to obtain satellite attitude data corresponding to multiple frames of continuous short-exposure images one by one, and storing the satellite attitude data into a third storage.
In the step, because the multi-frame continuous short-exposure images are shot in a motion state, the multi-frame continuous short-exposure images are not in one-to-one correspondence with the satellite attitude data, and therefore interpolation operation needs to be carried out on the satellite attitude data. The specific calculation process is as follows:
f(x)＝f(x0)+f[x0,x1](x-x0)+f[x0,x1,x2](x-x0)(x-x1)+…+f[x0,x1,…,xn-2,xn-1](x-x0)(x-x1)…(x-xn-2)(x-xn-1)+f[x0,x1,…,xn-1,xn](x-x0)(x-x1)…(x-xn-1)(x-xn)
wherein x is0、x1、x2、xn-2，xn-1、xnAs points on the function.
The attitude difference method mainly comprises three parts, wherein in the first step, through analyzing satellite attitude data generated by experimental data test, the satellite attitude data comprises three directions of pitching, rolling and rotating, and sinusoidal motion curves with different phases are respectively established.
The second part obtains satellite attitude data at the shooting moment through a Newton interpolation method, real attitude generation of a corresponding acquisition point of a multi-frame continuous short-exposure image is completed according to an original reference attitude by using bicubic interpolation, and the third part completes image correction according to shooting time satellite attitude data generated by noisy data.
And step S304, carrying out attitude calculation on the satellite attitude data after the interpolation calculation to obtain corrected data of multi-frame continuous short-exposure images which are in one-to-one correspondence with the attitude data after the interpolation calculation, and storing the corrected data into a fourth memory.
Specifically, the pose calculation is to translate and rotate the pixels in each of multiple frames of consecutive short-exposure images, for example, if the multiple frames of consecutive short-exposure images include N images, then N sets of correction data are obtained.
In step S305, pixel address mapping and image superposition are performed on pixels on multiple frames of continuous short-exposure images.
Specifically, the pixel address mapping is to calculate corresponding pixel address data according to the result of the gesture operation module. For example, for a 2560 × 2560 size image, 2560 pixels are in a row, the addresses of the 2560 pixels in the first row in the original image are calculated, the addresses of the 2560 pixels in the row are stored in the FIFO, after the 2560 addresses are stored, a read request is issued to the arbitration state machine, data is read out according to the row address and the column address, and meanwhile, the address of the next row is calculated, and the data is continuously written into the FIFO.
In the step, pixel address mapping continuously writes pixel address data into a pixel address FIFO at a speed of 50MHZ, applies for a request for reading the pixel address data when the number of the pixel address data in the FIFO is more than 2560, and READs out the data in the FIFO at a speed of 200MHZ and READs out 2560 data at a time if the READ state machine is entered. After a READ request is generated, a certain time may be required to enter the READ state (since the arbitration mechanism determines that the first memory is performing a READ/write operation, it is necessary to wait for the first memory to be idle), and in order to ensure that the pixel address FIFO does not overflow, the capacity of the FIFO is set to be 2 times of the number of pixels included in a row of pixels, thereby ensuring that the FIFO is not full. In addition, since the pixel address data is larger than 2560 before the read request is generated, the pixel address FIFO cannot be empty, thereby ensuring the reliability of the system.
In the step, in the image superposition process, when the superposition frame number is 9 frames, the final lifting multiple is about 2.5 times; when the number of superimposed frames is 16, the final lifting multiple is about 3.2 times.
It should be noted that the memory of the first memory includes a row address, a column address, and a bank address, and therefore, it is necessary to store pixel addresses in the row FIFO and the column FIFO, respectively. Meanwhile, the bit width of the user interface of the first memory is 128 bits, and the bit depth of the image is 16 bits, so bit width matching, namely address matching FIFO, is required to be performed, after completion, one line of pixel address data of the superposition cache region is read out and is placed into the line cache FIFO, and the steps are repeated until the superposition of all the multi-frame continuous short-exposure images is completed.
And step S306, transmitting the enhanced image data obtained by processing to any equipment for further processing, or converting the enhanced image data into a video signal and outputting the video signal to imaging equipment.
Specifically, the design of an application interface and interactive operation is completed, corresponding storage is needed for images shot by a camera and processed output images, the size of the images is stored as 10 bits, the size of display images in the system is stored as 8 bits, and the program has the capability of real-time demonstration.
Step S307, emptying data in the first memory, the second memory, the third memory, and the fourth memory.
Specifically, the enhanced image data is transmitted to video generation equipment and is reintegrated to show an infrared image which can be directly observed by human eyes, and the data in the memory is emptied, so that the enhancement processing of the next frame of image is facilitated.
Step S308, a real-time analysis is processed.
The time factor is an important factor for processing the image by the on-board FPGA, and the time consumed by the above steps S203 to S205 is described in detail below, and the image size is 2560 × 2560.
I. Interpolation operation stage
For satellite attitude data of 4 frames of images, 440ns is needed for interpolation operation; 460ns is needed for interpolation operation of satellite attitude data of 8 frames of images; the interpolation operation needs 490ns for satellite attitude data of 16 frames of images.
Attitude calculation stage
In the attitude operation stage, as seen from 195 system clock output results, 995ns is required for 4-frame image attitude operation, 1015ns is required for 8-frame image attitude operation, and 1055ns is required for 16-frame image attitude operation.
Address mapping and image overlay stage
Since the pixel address mapping calculation and the image correction superposition are carried out simultaneously, the pixel address mapping and the image superposition are carried out at the same time, the superposition time of 4 frames of images is 494ms, the superposition time of 8 frames of images is 986ms, and the superposition time of 16 frames of images is 1971 ms.
Example four
Fig. 4 is a schematic structural diagram of a system for processing a signal-to-noise ratio of a short-exposure image sequence in real time based on a satellite-borne camera according to an embodiment of the present disclosure, and as shown in fig. 4, the present disclosure further provides a system for processing a signal-to-noise ratio of a short-exposure image sequence in real time based on a satellite-borne camera, where the system includes:
the acquisition module I401 is used for acquiring a plurality of frames of continuous short-exposure images through the satellite-borne FPGA and storing the plurality of frames of continuous short-exposure images into a first memory;
the second acquisition module 402 is used for acquiring a plurality of satellite attitude data corresponding to the acquisition of a plurality of frames of continuous short-exposure images within shooting time through the satellite-borne FPGA control attitude sensor and storing the plurality of satellite attitude data into a second memory;
the first operation module 403 is used for performing interpolation operation on the satellite attitude data to obtain satellite attitude data corresponding to multiple frames of continuous short-exposure images one by one, and storing the satellite attitude data into a third memory;
the second operation module 404 is configured to perform attitude operation on the satellite attitude data after the interpolation operation to obtain corrected data of multiple frames of continuous short-exposure images, which correspond to the attitude data one by one, and store the corrected data in a fourth memory;
the third operation module 405 is configured to perform pixel address mapping calculation on pixels on multiple frames of continuous short-exposure images corresponding to the correction data one to one, and read pixel address data until all pixel address data on the multiple frames of continuous short-exposure images are read, so as to obtain enhanced image data;
the application module 406 is configured to transmit the enhanced image data obtained by the processing to any device for further processing, or convert the enhanced image data into a video signal and output the video signal to an imaging device;
and the processing module 407 is configured to clear data in the first memory, the second memory, the third memory, and the fourth memory.
Specifically, a plurality of frames of continuous short-exposure images are acquired through a first acquisition module 401, a plurality of satellite attitude data are acquired through a second acquisition module 402, interpolation operation is performed on the plurality of satellite attitude data through a first operation module 403, satellite attitude data corresponding to the plurality of frames of continuous short-exposure images one to one are acquired, attitude operation is performed through a second operation module 404, and then pixel address mapping and image superposition operation are performed on pixels on the plurality of frames of continuous short-exposure images to acquire and apply an enhanced image.
It should be noted that, when the power consumption of the system is 14.37W, it takes 986 milliseconds after 8 frames of images are superimposed, the system occupies 81% of on-chip storage space, 43.95% of off-chip storage space and 93.75% of multiplier resources of an FPGA with an XC5V L X110T model after the system is completed, and the weight of a hardware system platform is 0.47 kg.
According to the satellite-borne camera-based short-exposure image sequence signal-to-noise ratio real-time processing system, the programmable logic FPGA has great flexibility, image processing is performed by using the FPGA alone, the difficulty of pcb design and clock design is reduced, the FPGA of the sailing V5 series has a DSP48E operation hardmac, floating point number operation can be completed well, and the difficulty of floating point number operation is reduced to a certain extent.
In summary, the signal-to-noise ratio FPGA real-time processing module for the short-exposure image sequence of the satellite-borne camera provided by the invention at least achieves the following beneficial effects:
1. in the method and the system for processing the signal-to-noise ratio of the satellite-borne camera short-exposure image sequence in real time, a plurality of frames of continuous short-exposure images are acquired through a satellite-borne FPGA, a plurality of satellite attitude data are acquired through an attitude sensor controlled by the satellite-borne FPGA, interpolation operation is carried out on the satellite attitude data, the satellite attitude data corresponding to the plurality of frames of continuous short-exposure images one by one is acquired, the change of the satellite attitude data caused by the shake of shooting time can be eliminated, and the stability under the abnormal state is improved.
2. In the method and the system module for processing the signal-to-noise ratio of the short-exposure image sequence of the satellite-borne camera in real time, the signal-to-noise ratio of the image can be effectively improved through the superposition of multi-frame continuous short-exposure images, and the signal-to-noise ratio improvement multiple is increased along with the increase of the superposed frame number, but due to the registration deviation of the image, the signal-to-noise ratio improvement multiple does not completely meet the requirement of the signal-
The law of the multiplication is reduced to a certain extent.
3. According to the method and the system for processing the signal-to-noise ratio of the short-exposure image sequence of the satellite-borne camera in real time, the first storage, the second storage, the third storage and the fourth storage are called to store image data in the image processing process, and a pipeline type processing mode is used for the image processing process, so that the multi-step synchronous processing is realized, the image calculation speed can be greatly increased, and the hardware programming complexity and the memory space occupied by the system are reduced.
4. According to the method and the system for processing the signal-to-noise ratio of the short-exposure image sequence of the satellite-borne camera in real time, a hardware system based on an FPGA technology has great flexibility, and the difficulty of pcb design and clock design is reduced by using the FPGA for image processing.
Although some specific embodiments of the present invention have been described in detail by way of examples, it should be understood by those skilled in the art that the above examples are for illustrative purposes only and are not intended to limit the scope of the present invention. It will be appreciated by those skilled in the art that modifications may be made to the above embodiments without departing from the scope and spirit of the invention. The scope of the invention is defined by the appended claims.
Claims (10)
1. A signal-to-noise ratio real-time processing method for a short-exposure image sequence of a satellite-borne camera is characterized by comprising the following steps:
acquiring a plurality of frames of continuous short-exposure images through a satellite-borne FPGA, and storing the plurality of frames of continuous short-exposure images into a first memory;
controlling an attitude sensor to acquire a plurality of satellite attitude data corresponding to the multi-frame continuous short-exposure image shooting time through a satellite-borne FPGA (field programmable gate array), and storing the satellite attitude data into a second memory;
carrying out interpolation operation on the satellite attitude data to obtain satellite attitude data corresponding to the multiple frames of continuous short-exposure images one by one, and storing the satellite attitude data into a third memory;
carrying out attitude operation on the satellite attitude data after interpolation operation to obtain correction data of the multi-frame continuous short-exposure images, which are in one-to-one correspondence with the attitude data after interpolation operation, and storing the correction data into a fourth memory;
carrying out pixel address mapping calculation on pixels on the multiple frames of continuous short-exposure images which are in one-to-one correspondence with the correction data, and reading pixel address data until all the pixel address data on the multiple frames of continuous short-exposure images are read, so as to obtain enhanced image data;
transmitting the enhanced image data obtained by processing to any equipment for further processing or converting the enhanced image data into a video signal and outputting the video signal to imaging equipment;
and clearing the data in the first memory, the second memory, the third memory and the fourth memory.
2. The signal-to-noise ratio real-time processing method of the short-exposure image sequence of the satellite-borne camera according to claim 1, wherein the operation method of the interpolation operation is as follows: and inputting a horizontal and vertical coordinate of the attitude data and an interpolation enabling signal to the FPGA by adopting a secondary interpolation mode, and performing corresponding double-precision floating point number multiplication and division on the horizontal and vertical coordinate of the input attitude data and the interpolation enabling signal by the FPGA by adopting a floating point number operation IP core of the Sailing technology.
3. The signal-to-noise ratio real-time processing method of the short-exposure image sequence of the satellite-borne camera according to claim 1, wherein the attitude operation method comprises the following steps: inputting a plurality of parameters to the FPGA, wherein the parameters comprise corresponding satellite-borne FPGA shooting angles, attitude data, Gaussian templates and enabling signals, and performing operation by adopting a double-precision floating point type operation method.
4. The method for processing the signal-to-noise ratio of the short-exposure image sequence of the satellite-borne camera in real time according to claim 3, wherein the input of the corresponding parameters and the output of the corresponding results of the attitude calculation are performed in a pipeline mode.
5. The method for processing the signal-to-noise ratio of the short-exposure image sequence of the satellite-borne camera in real time according to claim 1, wherein the correction data comprises a translation matrix and a rotation matrix.
6. The method for processing the signal-to-noise ratio of the short-exposure image sequence of the satellite-borne camera in real time according to claim 1, wherein the pixel address mapping calculation is performed on the pixels of the multiple frames of continuous short-exposure images which are in one-to-one correspondence with the correction data, and the pixel address data is read until all the pixel address data of the multiple frames of continuous short-exposure images are read, so that the enhanced image data is obtained, and specifically:
and sequentially carrying out pixel address mapping calculation on each line of pixels in the multiple frames of continuous short-exposure images in one-to-one correspondence to the correction data, reading one line of pixel address data after completing one line of pixel address mapping calculation, carrying out superposition operation, and completing the whole image calculation until all the pixel address data on the multiple frames of continuous short-exposure images in one-to-one correspondence to the correction data are completed, so as to obtain enhanced image data.
7. The method for processing the signal-to-noise ratio of the short-exposure image sequence of the satellite-borne camera in real time according to claim 6, wherein the capacity of the pixel address is 2 times the number of pixels contained in one row of pixels.
8. The method for processing the signal-to-noise ratio of the short-exposure image sequence of the satellite-borne camera in real time according to claim 1, wherein before the enhanced image data obtained by processing is transmitted to other equipment for further processing or is converted into a video signal and output to imaging equipment for display, the method comprises the following steps:
and storing the enhanced image obtained by processing, wherein the size of the image storage is 10 bits.
9. The method for processing the signal-to-noise ratio of the short-exposure image sequence of the satellite-borne camera in real time according to claim 1, wherein the attitude data comprises three-dimensional space information and time information, and the three-dimensional space information comprises a pitch angle, a translation angle and a turning angle.
10. A signal-to-noise ratio real-time processing system based on a short-exposure image sequence of a satellite-borne camera is characterized by comprising:
the acquisition module I is used for acquiring a plurality of frames of continuous short-exposure images through a satellite-borne FPGA and storing the plurality of frames of continuous short-exposure images into a first memory;
the acquisition module II is used for acquiring a plurality of satellite attitude data corresponding to the acquisition of the multi-frame continuous short-exposure image shooting time through a satellite-borne FPGA control attitude sensor and storing the satellite attitude data into a second memory;
the first operation module is used for carrying out interpolation operation on the satellite attitude data to obtain satellite attitude data corresponding to the multiple frames of continuous short-exposure images one by one and storing the satellite attitude data into a third memory;
the second operation module is used for carrying out attitude operation on the satellite attitude data after interpolation operation to obtain correction data of the multiple frames of continuous short-exposure images, which are in one-to-one correspondence with the attitude data, and storing the correction data into a fourth memory;
the operation module III is used for carrying out pixel address mapping calculation on the pixels on the multiple frames of continuous short-exposure images which are in one-to-one correspondence with the correction data, and reading pixel address data until all the pixel address data on the multiple frames of continuous short-exposure images are read, so as to obtain enhanced image data;
the application module is used for transmitting the enhanced image data obtained by processing to any equipment for further processing or converting the enhanced image data into a video signal and outputting the video signal to the imaging equipment;
and the processing module is used for emptying the data in the first memory, the second memory, the third memory and the fourth memory.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010311683.8A CN111464755B (en) | 2020-04-20 | 2020-04-20 | Signal-to-noise ratio real-time processing method and system for short-exposure image sequence of satellite-borne camera |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010311683.8A CN111464755B (en) | 2020-04-20 | 2020-04-20 | Signal-to-noise ratio real-time processing method and system for short-exposure image sequence of satellite-borne camera |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111464755A true CN111464755A (en) | 2020-07-28 |
| CN111464755B CN111464755B (en) | 2021-06-18 |
Family
ID=71682566
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010311683.8A Active CN111464755B (en) | 2020-04-20 | 2020-04-20 | Signal-to-noise ratio real-time processing method and system for short-exposure image sequence of satellite-borne camera |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111464755B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102044053A (en) * | 2009-10-20 | 2011-05-04 | 西安费斯达自动化工程有限公司 | Inverse perspective mapping (IPM) method based on field programmable gate array (FPGA) |
| CN103248820A (en) * | 2012-02-10 | 2013-08-14 | 佳能株式会社 | Image pickup apparatus and control method |
| TWM501362U (en) * | 2014-11-14 | 2015-05-21 | Shui-Mu Wang | Improved structure of the composite material |
| CN106027904A (en) * | 2016-06-29 | 2016-10-12 | 中国人民解放军国防科学技术大学 | Video satellite attitude motion image compensation method |
| CN106197402A (en) * | 2016-07-04 | 2016-12-07 | 北京航空航天大学 | A kind of multiple-exposure formation method of image intensifying type star sensor |
| CN108605098A (en) * | 2016-05-20 | 2018-09-28 | 深圳市大疆创新科技有限公司 | system and method for rolling shutter correction |
| CN109360229A (en) * | 2018-10-31 | 2019-02-19 | 歌尔股份有限公司 | Laser projection image processing method, device and equipment |
| CN109410130A (en) * | 2018-09-28 | 2019-03-01 | 华为技术有限公司 | Image processing method and image processing apparatus |
| WO2019118319A1 (en) * | 2017-12-15 | 2019-06-20 | Gopro, Inc. | High dynamic range processing on spherical images |
-
2020
- 2020-04-20 CN CN202010311683.8A patent/CN111464755B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102044053A (en) * | 2009-10-20 | 2011-05-04 | 西安费斯达自动化工程有限公司 | Inverse perspective mapping (IPM) method based on field programmable gate array (FPGA) |
| CN103248820A (en) * | 2012-02-10 | 2013-08-14 | 佳能株式会社 | Image pickup apparatus and control method |
| TWM501362U (en) * | 2014-11-14 | 2015-05-21 | Shui-Mu Wang | Improved structure of the composite material |
| CN108605098A (en) * | 2016-05-20 | 2018-09-28 | 深圳市大疆创新科技有限公司 | system and method for rolling shutter correction |
| CN106027904A (en) * | 2016-06-29 | 2016-10-12 | 中国人民解放军国防科学技术大学 | Video satellite attitude motion image compensation method |
| CN106197402A (en) * | 2016-07-04 | 2016-12-07 | 北京航空航天大学 | A kind of multiple-exposure formation method of image intensifying type star sensor |
| WO2019118319A1 (en) * | 2017-12-15 | 2019-06-20 | Gopro, Inc. | High dynamic range processing on spherical images |
| CN109410130A (en) * | 2018-09-28 | 2019-03-01 | 华为技术有限公司 | Image processing method and image processing apparatus |
| CN109360229A (en) * | 2018-10-31 | 2019-02-19 | 歌尔股份有限公司 | Laser projection image processing method, device and equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111464755B (en) | 2021-06-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108898630B (en) | Three-dimensional reconstruction method, device, equipment and storage medium | |
| KR101939349B1 (en) | Around view video providing method based on machine learning for vehicle | |
| CN109618090B (en) | Method and system for image distortion correction of images captured using wide angle lenses | |
| JP3058774B2 (en) | Image synthesizing apparatus and image synthesizing method | |
| CN110909790A (en) | Image style migration method, device, terminal and storage medium | |
| CN112837245B (en) | Dynamic scene deblurring method based on multi-mode fusion | |
| US20140078347A1 (en) | Systems and Methods for Reducing Noise in Video Streams | |
| CN111199528A (en) | Fisheye image distortion correction method | |
| CN104284126A (en) | Image interpolation method, image interpolation device and image device using same | |
| CN105981073A (en) | Image generating apparatus, image generating method, and program | |
| CN113315884A (en) | Real-time video noise reduction method and device, terminal and storage medium | |
| CN114463230A (en) | Image processing method and related equipment | |
| CN110211039B (en) | Image processing method and device | |
| Song et al. | Real-scene reflection removal with raw-rgb image pairs | |
| CN111464755B (en) | Signal-to-noise ratio real-time processing method and system for short-exposure image sequence of satellite-borne camera | |
| CN107707820A (en) | Aerial camera real-time electronic zooming system based on FPGA | |
| JP6113406B2 (en) | Infrared target detection device | |
| CN115937374A (en) | Digital human body modeling method, device, equipment and medium | |
| JPH1074258A (en) | Image processing method and device therefor | |
| CN116017123A (en) | Photoelectric imaging system and imaging method thereof | |
| JP5682473B2 (en) | Image processing method having wide-angle distortion correction processing, image processing apparatus, and imaging apparatus | |
| CN115546373A (en) | NeRF renderer based on FPGA and multi-resolution Hash coding algorithm and rendering method | |
| JP2013005393A (en) | Image processing method having wide-angle distortion correction processing, image processing apparatus and imaging apparatus | |
| Angelopoulou et al. | Robust real-time super-resolution on FPGA and an application to video enhancement | |
| CN114820299A (en) | Non-uniform motion blur super-resolution image restoration method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |