High-speed image two-way real-time frame extraction resolving method

Abstract

The invention relates to a high-speed image two-way real-time frame extraction resolving method, which comprises the following steps of: setting imaging parameters, transmitting the imaging parameters to a control panel through a 422 interface, and transmitting the imaging parameters to a high-speed focal plane through the 422 interface by the control panel; the high-speed focal plane receives the imaging parameter instruction to perform high-speed imaging, generates an M-frame / second image, and performs unique numbering on the M-frame / second image; and performing frame extraction calculation on the generated M frames / second image to obtain an image containing a target and target information. According to the invention, real-time processing of a high-speed image can be realized on a non-ground platform, information of a high-speed moving object in the image is obtained, and the method is used for real-time monitoring of the high-speed moving object by an external field high-altitude platform, and belongs to the technical field of real-time processing of large-data-volume images.

Description

Technical Field

[0001] The invention relates to a high-speed image dual-path real-time frame extraction and solving method, belonging to the technical field of real-time processing of large-volume images. Background Art

[0002] With the continuous development of high-speed imaging technology, the imaging frame rate is getting higher and higher, the array size is getting larger and larger, and the image quantization bit is getting larger and larger, so the requirements for real-time storage and real-time processing of large amounts of image data are also getting higher and higher. The real-time storage and processing of high-speed images on non-ground platforms will be subject to the volume, weight, power consumption and other requirements of the platform. The existing technology is difficult to meet the application requirements, and a new high-speed image real-time solution method is needed to meet the needs of real-time storage and processing.

[0003] In terms of real-time storage of large amounts of high-speed images, traditional methods mainly include DDR cache, Flash storage, mechanical hard disk storage, SSD storage, etc. Traditional DDR cache has a high storage rate, but limited storage capacity; Flash image data acquisition and storage, complex interface operation, poor stability; mechanical hard disk storage capacity is small and slow; SSD storage is much faster than traditional hard disks, more durable, low power consumption, large storage capacity, a single chip can achieve 1T, 2T, 4T, etc., can achieve a write rate of 6000-7000MB / s for a long time.

[0004] In terms of real-time processing of large numbers of high-speed images, traditional methods mainly include workstations, PCs, etc. Workstations have fast processing speeds and can use commercial software to enhance the target flow field processing capabilities, but the system is complex, large in size, heavy in weight, and is generally located on the ground, and is not suitable for non-ground platforms; PC processing capabilities are slightly worse than workstations, and there are also problems with complex processing systems, large in size, and heavy in weight. It is generally located on the ground and is not suitable for non-ground platforms. Summary of the invention

[0005] The technical problem solved by the present invention is: to overcome the shortcomings of the prior art and propose a high-speed processing system for target flow fields with large amounts of data, so as to solve the difficult problems of insufficient high-speed data transmission capacity, insufficient high-speed data storage capacity and insufficient high-speed data processing capacity in the existing high-speed processing systems in the target flow field processing tasks of air-based and space-based platforms.

[0006] The technical solution of the present invention is: a high-speed image dual-path real-time frame extraction and solution method, comprising:

[0007] The user sets the imaging parameters through the control interface of the display control board, and the imaging parameters are transmitted to the control board through the 422 interface. The control board transmits the imaging parameters to the high-speed focal plane through the 422 interface;

[0008] The high-speed focal plane receives an imaging parameter instruction to perform high-speed imaging, generates M frames / second images, and uniquely numbers the M frames / second images;

[0009] The generated M frames / second image is subjected to frame extraction and decoding to obtain an image containing the target and target information.

[0010] The M frames / second images are uniquely numbered, and the M frames / second image numbering rule is the clock information year, month, day, hour, minute, second and image sequence number.

[0011] The generated M frames / second images are respectively extracted and solved through two paths, namely, through a Gigabit or 10 Gigabit network port and an Aurora high-speed data transmission interface, to finally obtain an image containing the target and target information.

[0012] Frame extraction and analysis through Gigabit or 10 Gigabit network ports, including:

[0013] Through the Gigabit or 10 Gigabit network port, N frames / second images are extracted from M frames / second images and transmitted to the processing board;

[0014] The processing board processes the N frames / second image in real time, identifies the image containing the target, extracts the frame number of the image, and transmits the image frame number and target information containing the target to the control board through the serial port;

[0015] The control board transmits the image frame number containing the target to the high-speed SSD through the serial port, and extracts the image corresponding to the image frame number containing the target from the high-speed SSD through the Aurora high-speed data transmission interface as the image with a specific frame number;

[0016] The control panel transmits the specific frame number image and target information to the display control panel in real time through the Gigabit network, the 10 Gigabit network or the optical fiber, and the display control panel displays the specific frame number image and target information including the target.

[0017] The N-frame / second image is extracted from the M-frame / second image, and the extraction method adopts one of the following three methods:

[0018] Method 1: single frames are drawn at equal intervals, i.e., 0 frame images, M / N frame images, 2M / N...N*M / N frame images per second;

[0019] Method 2: continuously extract K frames of images at equal intervals, that is, 0 frame, 1 frame, 2 frame...K frame images per second, M / (N / K) frame, M / (N / K)+1 frame, M / (N / K)+2 frame...M / (N / K)+K frame images, 2M / (N / K) frame, 2M / (N / K)+1 frame, 2M / (N / K)+2 frame...2M / (N / K)+K frame images,...;

[0020] Method 3: Randomly extract frames, that is, randomly extract frames through a random function; the N value must meet the following requirements:

[0021]

[0022] DK is the available bandwidth of Gigabit network, 10 Gigabit network or optical fiber interface; P is the size of a single frame image.

[0023] The processing board has a built-in image processing program for real-time processing of N frames / second images; the image processing program can be updated through a script file, and the user inputs the update command through the display control board, and the update command is transmitted to the control board, and then transmitted to the processing board through the control board.

[0024] Frame extraction and analysis are performed through the Aurora high-speed data transmission interface, including:

[0025] Through the Aurora high-speed data transmission interface, all M frames / second images are transmitted to DDR for caching, and DDR then stores the images in the high-speed SSD;

[0026] The user inputs the frame extraction instruction according to the application requirements through the control interface of the display control board, and transmits the frame extraction instruction to the control board through the 422 interface;

[0027] The control board extracts the images required by the user from the SSD in real time through the Aurora high-speed data transmission interface according to the frame extraction instructions specified by the user;

[0028] The user transmits the required images to the display control board in real time through Gigabit network, 10 Gigabit network or optical fiber;

[0029] The display control panel processes the images required by the user in real time and identifies the images and target information containing the target.

[0030] The method of transmitting all M frames / second images to DDR for caching, and DDR then storing the images in a high-speed SSD includes: dividing the M frames / second images into single-numbered and double-numbered images, storing them in DDR1 and DDR2 respectively, and then storing the images in DDR1 in high-speed SSD1 and the images in DDR2 in high-speed SSD2, thereby improving the real-time storage rate and realizing real-time storage of M frames / second images.

[0031] All images stored in SSD1 and SSD2 are cyclically overwritten when the storage capacity is full; all images stored in SSD1 and SSD2 are exported to the display control panel or the ground system through the full data export interface after step twelve.

[0032] The display control board transmits the frame extraction instruction to the control board through the 422 interface, wherein the frame extraction instruction has three modes to choose from, namely, extracting a single frame at equal intervals, continuously extracting K frames of images at equal intervals, and randomly extracting frames.

[0033] The advantages of the present invention compared with the prior art are:

[0034] 1) The present invention adopts a Gigabit or 10 Gigabit network interface between the high-speed focal plane and the processing board to extract N frames / second images from M frames / second high-speed images and transmit them to the processing board in real time. Through the built-in image processing program of the processing board, real-time frame extraction and resolution of one high-speed image is realized, which can effectively solve the problem of real-time processing of high-speed images.

[0035] 2) The present invention adopts a method in which the display control board transmits the user frame extraction instruction to the control board, and the control board extracts the corresponding image from the high-speed SSD. The image is transmitted from the control board to the display control board, thereby realizing another high-speed image real-time frame extraction and solution, which can effectively solve the problem of real-time processing of high-speed images.

[0036] 3) The present invention adopts a method of storing odd and even number images in high-speed imaging respectively through dual-channel DDR + high-speed SSD, so as to realize real-time storage of a large number of images with large array, high frame rate and high quantization bit number, and solve the problem that large amount of high-speed images cannot be stored in real time.

[0037] 4) The present invention exports all images stored in SSD1 and SSD2 to the display control panel or the ground system through the full data export interface after step twelve, which can not only meet the working mode requirements of the existing traditional high-speed imaging system, but also meet the requirements of the dual-channel real-time frame extraction and solution mode, and can meet the ground and non-ground usage requirements at the same time.

[0038] 5) The present invention adopts three methods: single frame extraction at equal intervals, continuous extraction of K frames at equal intervals, and random frame extraction to solve the problem of missing objects in the image when imaging high-speed moving objects in the fixed frame extraction method, and can also minimize the impact of N frames / second frame extraction on the original data M frames / second high-speed imaging.

[0039] 6) The image processing program built into the processing board of the present invention can be updated through script files to meet the needs of different users. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 The present invention is a schematic diagram of the composition of a high-speed image dual-path real-time frame extraction and solution method. DETAILED DESCRIPTION

[0041] like Figure 1As shown, the present invention uses a Gigabit or 10 Gigabit network interface between the high-speed focal plane and the processing board to extract N frames / second images from M frames / second high-speed images, and transmit them to the processing board in real time. Through the built-in image processing program of the processing board, real-time frame extraction and resolution of one channel of high-speed images is realized. The present invention uses a method in which the display control board transmits the user's frame extraction instruction to the control board, the control board extracts the corresponding image from the high-speed SSD, and the image is transmitted from the control board to the display control board, thereby realizing real-time frame extraction and resolution of another channel of high-speed images. The dual-channel real-time frame extraction and resolution method can be performed simultaneously or separately. The specific steps are as follows:

[0042] The user sets the imaging parameters through the control interface of the display control board, and the imaging parameters are transmitted to the control board through the 422 interface. The control board transmits the imaging parameters to the high-speed focal plane through the 422 interface;

[0043] The high-speed focal plane receives an imaging parameter instruction to perform high-speed imaging, generates M frames / second images, and uniquely numbers the M frames / second images;

[0044] The generated M frames / second image is subjected to frame extraction and decoding to obtain an image containing the target and target information.

[0045] The M frames / second images are uniquely numbered, and the M frames / second image numbering rule is the clock information year, month, day, hour, minute, second and image sequence number.

[0046] The generated M frames / second images are respectively extracted and solved through two paths, namely, through a Gigabit or 10 Gigabit network port and an Aurora high-speed data transmission interface, to finally obtain an image containing the target and target information.

[0047] Frame extraction and analysis through Gigabit or 10 Gigabit network ports, including:

[0048] Through the Gigabit or 10 Gigabit network port, N frames / second images are extracted from M frames / second images and transmitted to the processing board;

[0049] The processing board processes the N frames / second image in real time, identifies the image containing the target, extracts the frame number of the image, and transmits the image frame number and target information containing the target to the control board through the serial port;

[0050] The control board transmits the image frame number containing the target to the high-speed SSD through the serial port, and extracts the image corresponding to the image frame number containing the target from the high-speed SSD through the Aurora high-speed data transmission interface as the image with a specific frame number;

[0051] The control panel transmits the specific frame number image and target information to the display control panel in real time through the Gigabit network, the 10 Gigabit network or the optical fiber, and the display control panel displays the specific frame number image and target information including the target.

[0052] The N-frame / second image is extracted from the M-frame / second image, and the extraction method adopts one of the following three methods:

[0053] Method 1: single frames are drawn at equal intervals, i.e., 0 frame images, M / N frame images, 2M / N...N*M / N frame images per second;

[0054] Method 2: continuously extract K frames of images at equal intervals, that is, 0 frame, 1 frame, 2 frame...K frame images per second, M / (N / K) frame, M / (N / K)+1 frame, M / (N / K)+2 frame...M / (N / K)+K frame images, 2M / (N / K) frame, 2M / (N / K)+1 frame, 2M / (N / K)+2 frame...2M / (N / K)+K frame images,...;

[0055] Method 3: Randomly extract frames, that is, randomly extract frames through a random function; the N value must meet the following requirements:

[0056]

[0057] DK is the available bandwidth of Gigabit network, 10 Gigabit network or optical fiber interface; P is the size of a single frame image.

[0058] The processing board has a built-in image processing program for real-time processing of N frames / second images; the image processing program can be updated through a script file, and the user inputs the update command through the display control board, and the update command is transmitted to the control board, and then transmitted to the processing board through the control board.

[0059] Frame extraction and analysis are performed through the Aurora high-speed data transmission interface, including:

[0060] Through the Aurora high-speed data transmission interface, all M frames / second images are transmitted to DDR for caching, and DDR then stores the images in the high-speed SSD;

[0061] The user inputs the frame extraction instruction according to the application requirements through the control interface of the display control board, and transmits the frame extraction instruction to the control board through the 422 interface;

[0062] The control board extracts the images required by the user from the SSD in real time through the Aurora high-speed data transmission interface according to the frame extraction instructions specified by the user;

[0063] The user transmits the required images to the display control board in real time through Gigabit network, 10 Gigabit network or optical fiber;

[0064] The display control panel processes the images required by the user in real time and identifies the images and target information containing the target.

[0065] The method of transmitting all M frames / second images to DDR for caching, and DDR then storing the images in a high-speed SSD includes: dividing the M frames / second images into single-numbered and double-numbered images, storing them in DDR1 and DDR2 respectively, and then storing the images in DDR1 in high-speed SSD1 and the images in DDR2 in high-speed SSD2, thereby improving the real-time storage rate and realizing real-time storage of M frames / second images.

[0066] All images stored in SSD1 and SSD2 are cyclically overwritten when the storage capacity is full; all images stored in SSD1 and SSD2 are exported to the display control panel or the ground system through the full data export interface after step twelve.

[0067] The display control board transmits the frame extraction instruction to the control board through the 422 interface, wherein the frame extraction instruction has three modes to choose from, namely, extracting a single frame at equal intervals, continuously extracting K frames of images at equal intervals, and randomly extracting frames.

[0068] The images obtained from high-speed focal plane imaging in the present invention are all stored in DDR and then stored in a high-speed SSD, wherein DDR can be DDR3, DDR4, DDR5, DDR6, etc., and can be selected according to the array size, frame rate, quantization bit number, etc. of the high-speed imaging.

[0069] The image processing program built into the processing board of the present invention can be set as a default program by the user according to needs, or updated through a script program. The image processing program is not limited to the detection of motion information of high-speed moving objects, and other information required by users can also be obtained by running the image processing program.

[0070] The image obtained by high-speed focal plane imaging in the present invention extracts N frames / second images from M frames / second images through a data interface and transmits them to a processing board. The interface here can be a gigabit or 10 gigabit network port, or it can be other types of interfaces, as long as the interface requirements of the focal plane and the processing board are met. There are three ways of extracting frames here, namely, extracting single frames at equal intervals, continuously extracting K frames of images at equal intervals, and randomly extracting frames, wherein the equal interval Q of the single frame at equal intervals can be specified by the user, and the equal interval Q can be divided by M or not by M, and the frame extraction sequence starts from the image No. 0 of the image per second, but the total image size of the frame extraction cannot exceed the interface bandwidth, and the image exceeding the interface bandwidth cannot be transmitted.

[0071] In the present invention, the user inputs a new frame extraction instruction through the control interface of the display control board. Here, the new frame extraction instruction has three modes, namely, single frame extraction at equal intervals, continuous extraction of K frames at equal intervals, and random frame extraction. The total image size of the frame extraction cannot exceed the interface bandwidth. The three frame extraction modes here are independent of the three frame extraction modes from high-speed focal plane imaging to the processing board.

[0072] The present invention can also embed a processing program in the display control board, and the processing program is determined by user needs.

[0073] The present invention can partition SSD1 and SSD2, and after partitioning, the storage area can be specified, and the image stored in a partition will not be overwritten before it is selected to be stored again.

[0074] All images stored in SSD1 and SSD2 of the present invention can be exported to the display control board or the ground system through the interface afterwards. The transmission interface can be fast export or slow export, which is determined by user needs.

[0075] Although the present invention has been disclosed as above with preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make possible changes and modifications to the technical solution of the present invention by using the above disclosed methods and technical contents without departing from the spirit and scope of the present invention. Therefore, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present invention without departing from the content of the technical solution of the present invention shall belong to the protection scope of the technical solution of the present invention. In the absence of conflict, the embodiments of the present application and the technical features in the embodiments can be combined with each other.

Claims

1. A high-speed image dual-channel real-time frame extraction and solution method, characterized in that: include: Setting imaging parameters, wherein the imaging parameters are transmitted to the control board through the 422 interface, and the control board transmits the imaging parameters to the high-speed focal plane through the 422 interface; The high-speed focal plane receives an imaging parameter instruction to perform high-speed imaging, generates M frames / second images, and uniquely numbers the M frames / second images; The generated M frames / second image is subjected to frame extraction and decoding to obtain an image containing the target and target information.

2. A high-speed image dual-path real-time frame extraction and solving method according to claim 1, characterized in that: The M frames / second images are uniquely numbered, and the M frames / second image numbering rule is the clock information year, month, day, hour, minute, second and the image sequence number.

3. The high-speed image dual-path real-time frame extraction and solving method according to claim 1 is characterized by: The generated M frames / second images are respectively extracted and solved through two paths, namely, through a Gigabit or 10 Gigabit network port and an Aurora high-speed data transmission interface, to finally obtain an image containing the target and target information.

4. A high-speed image dual-channel real-time frame extraction and solving method according to claim 3, characterized in that: Frame extraction and analysis through Gigabit or 10 Gigabit network ports, including: Through the Gigabit or 10 Gigabit network port, N frames / second images are extracted from M frames / second images and transmitted to the processing board; The processing board processes the N frames / second image in real time, identifies the image containing the target, extracts the frame number of the image, and transmits the image frame number and target information containing the target to the control board through the serial port; The control board transmits the image frame number containing the target to the high-speed SSD through the serial port, and extracts the image corresponding to the image frame number containing the target from the high-speed SSD through the Aurora high-speed data transmission interface as the image with a specific frame number; The control panel transmits the specific frame number image and target information to the display control panel in real time through the Gigabit network, the 10 Gigabit network or the optical fiber, and the display control panel displays the specific frame number image and target information including the target.

5. A high-speed image dual-channel real-time frame extraction and solving method according to claim 4, characterized in that: The N-frame / second image is extracted from the M-frame / second image, and the extraction method adopts one of the following three methods: Method 1: single frames are drawn at equal intervals, i.e., 0 frame images, M / N frame images, 2M / N...N*M / N frame images per second; Method 2: continuously extract K frames of images at equal intervals, that is, 0 frame, 1 frame, 2 frame...K frame images per second, M / (N / K) frame, M / (N / K)+1 frame, M / (N / K)+2 frame...M / (N / K)+K frame images, 2M / (N / K) frame, 2M / (N / K)+1 frame, 2M / (N / K)+2 frame...2M / (N / K)+K frame images,...; Method 3: Randomly extract frames, that is, randomly extract frames through a random function; the N value must meet the following requirements: DK is the available bandwidth of Gigabit network, 10 Gigabit network or optical fiber interface; P is the size of a single frame image.

6. A high-speed image dual-path real-time frame extraction and solving method according to claim 4, characterized in that: The processing board has a built-in image processing program for real-time processing of N frames / second images; the image processing program can be updated through a script file, and the user inputs the update command through the display control board, and the update command is transmitted to the control board, and then transmitted to the processing board through the control board.

7. A high-speed image dual-path real-time frame extraction and solving method according to claim 5, characterized in that: Frame extraction and analysis are performed through the Aurora high-speed data transmission interface, including: Through the Aurora high-speed data transmission interface, all M frames / second images are transmitted to DDR for caching, and DDR then stores the images in the high-speed SSD; The user inputs the frame extraction instruction according to the application requirements through the control interface of the display control board, and transmits the frame extraction instruction to the control board through the 422 interface; The control board extracts the images required by the user from the SSD in real time through the Aurora high-speed data transmission interface according to the frame extraction instructions specified by the user; The user transmits the required images to the display control board in real time through Gigabit network, 10 Gigabit network or optical fiber; The display control panel processes the images required by the user in real time and identifies the images and target information containing the target.

8. A high-speed image dual-path real-time frame extraction and solving method according to claim 7, characterized in that: The method of transmitting all M frames / second images to DDR for caching, and DDR then storing the images in a high-speed SSD includes: dividing the M frames / second images into single-numbered and double-numbered images, storing them in DDR1 and DDR2 respectively, and then storing the images in DDR1 in high-speed SSD1 and the images in DDR2 in high-speed SSD2, thereby improving the real-time storage rate and realizing real-time storage of M frames / second images.

9. A high-speed image dual-path real-time frame extraction and solving method according to claim 8, characterized in that: All images stored in SSD1 and SSD2 will be overwritten cyclically when the storage capacity is full; All images stored in SSD1 and SSD2 are exported to the display control panel or the ground system through the full data export interface after step 12.

10. The high-speed image dual-path real-time frame extraction and solving method according to claim 7, characterized in that: The display control board transmits the frame extraction instruction to the control board through the 422 interface, wherein the frame extraction instruction has three modes to choose from, namely, extracting a single frame at equal intervals, continuously extracting K frames of images at equal intervals, and randomly extracting frames.

Images