Time division multiplexing and sliding template matching head-up display symbol recognition method and system

Abstract

The present invention relates to a time-division multiplexing and sliding template matching method and system for symbol recognition of a head-up display, including a timestamp recognition unit, an airspeed intercepting unit, an atmospheric pressure altitude intercepting unit, a metric altitude intercepting unit, a Mach number intercepting unit, an artificial horizon recognition unit, a roll angle calculation unit, a pitch ladder feature point recognition unit, a rotation calculation unit, a dual-port memory unit, a control center unit, a sliding template matching unit, a symbol template unit, and a result summary unit, which are used to solve the problem of recognizing six key symbols, namely airspeed, atmospheric pressure altitude, metric altitude, Mach number, roll angle, and pitch ladder, in the video image of the head-up display. On the one hand, the invention introduced by this method improves the integrity and real-time performance of the symbol recognition method through time-division multiplexing technology, saves computing resources, and reduces hardware costs. On the other hand, it improves the accuracy and precision of symbol recognition through sliding template matching technology.

Description

Technical Field

[0001] The present invention belongs to the technical field of video signal processing and image processing, and relates to a time-division multiplexing and sliding template matching head-up display symbol recognition method and system, which can be applied to the recognition and monitoring links of key symbols in the video images of airborne avionics equipment, and improve the safety of airborne avionics equipment. Background Art

[0002] In the aviation field, airborne avionics equipment has high safety requirements. To meet the safety of the head-up display, the head-up display system needs to effectively monitor the generated video images, determine whether the key symbols in the video images are consistent with the airborne avionics parameters, and if not, report them in time to avoid misleading the pilot. Accurately identifying the key symbols in the video images is the first and most important link in the monitoring link. The head-up display symbol recognition method needs to meet requirements such as high integrity, high real-time performance, high accuracy, and low false alarm rate. Traditional image recognition technologies cannot meet the performance requirements of head-up display symbol recognition. For example, excessive delay may affect flight safety. Currently in China, the head-up display symbol recognition method is still in its infancy. Summary of the Invention

[0003] Technical Problems to be Solved

[0004] To avoid the deficiencies of the prior art, the present invention proposes a time-division multiplexing and sliding template matching head-up display symbol recognition method and system, and provides a head-up display symbol recognition method with high integrity, high real-time performance, high accuracy, low false alarm rate, and low cost.

[0005] Technical Solution

[0006] A time-division multiplexing and sliding template matching head-up display symbol recognition method, characterized by the following steps:

[0007] Step 1: Receive the head-up display video image, respectively identify the timestamps in the video image, and at the same time read the green channel pixel values in the airspeed, barometric altitude, metric altitude, and Mach number regions of the video image, and complete the binarization operation to obtain the binarized images of the intercepted regions in the airspeed, barometric altitude, metric altitude, and Mach number regions;

[0008] At the same time, identify the white artificial horizon in the video image, detect the uppermost and lowermost white pixel points of the artificial horizon, and calculate the roll angle based on the coordinates of these two white pixel points;

[0009] At the same time, identify the pitch ladder feature points in the video image, detect the four blue feature points around the pitch ladder symbol in the video image, and obtain the coordinates of the four blue feature points;

[0010] Step 2: Determine the coordinate area of the pitching ladder based on the coordinates of four blue feature points, perform bilinear interpolation operation on the pitching ladder with the roll angle, rotate the pitching ladder to the horizontal state, and obtain the binary pixel values after rotation;

[0011] Step 3: According to the recognized timestamp, judge the parity of the timestamp;

[0012] When the timestamp is odd, first perform sliding template matching operation on the binary image of the airspeed intercept area, and then perform sliding template matching operation on the binary image of the Mach number intercept area;

[0013] When the timestamp is even, first perform sliding template matching operation on the binary image of the metric height intercept area, then perform sliding template matching operation on the binary image of the barometric height intercept area, and finally perform sliding template matching operation on the pitching ladder intercept area;

[0014] Step 4: After the sliding template matching operation is completed, all the symbols to be recognized in the current frame are matched, and the matching results of all symbols are obtained.

[0015] When it is necessary to match the next frame of image, repeat Steps 1 to 3.

[0016] The described sliding template matching method: First, overlap the key symbol area intercepted from the video frame with the upper left corner of the symbol template and perform template matching operation; after the operation is completed, move the intercepted area one column to the right and perform template matching operation; after the operation is completed, continue to move the intercepted area one column to the right until the rightmost column of the intercepted area coincides with the rightmost side of the symbol template and perform template matching operation; after the operation is completed, overlap the intercepted area with the left side of the symbol template and move down one row to perform template matching operation until the intercepted area coincides with the lower right corner of the symbol template, and the sliding template matching operation ends.

[0017] During the process of the video intercept area moving from the upper left corner to the lower right corner of the symbol template, for each template matching result, find the maximum value of the template matching result and the template area corresponding to the maximum value, and the template number in this area is the sliding template matching result of the intercepted area.

[0018] The described symbol template needs to cover all situations where the key symbols appear in the video frame, ensuring that the key symbol intercept area of any frame can find a substantially overlapping area in the corresponding symbol template.

[0019] The acquisition of the roll angle: For the white pixel points at the top and bottom of the detected artificial horizon, calculate the roll angle through the arctangent function.

[0020] The pitch ladder extraction area: Extract the lower left corner feature point among the four feature points around the pitch ladder, use this feature point as the rotation center for rotation calculation, and at the same time use the detected uppermost feature point as the reference point of the extraction area. With the abscissa of this reference point as the center, expand 35 pixels to the left and right respectively, and use the ordinate of the reference point as the upper boundary of the extraction area, and expand 42 pixels downward. The obtained 42×70 area is the pitch ladder extraction area.

[0021] A system for implementing the above-mentioned head-up display symbol recognition method based on time-division multiplexing and sliding template matching, characterized by including a timestamp recognition unit, an airspeed extraction unit, a barometric altitude extraction unit, a metric altitude extraction unit, a Mach number extraction unit, an artificial horizon recognition unit, a roll angle calculation unit, a pitch ladder feature point recognition unit, a rotation calculation recognition unit, a dual-port memory unit, a control center unit, a sliding template matching unit, a symbol template unit, and a result summary unit; The timestamp recognition unit, the airspeed extraction unit, the barometric altitude extraction unit, the metric altitude extraction unit, the Mach number extraction unit, and the artificial horizon recognition unit simultaneously receive the input image information, where

[0022] The timestamp recognition unit recognizes the timestamp in the video frame, judges the parity of the timestamp, and sends the timestamp recognition result to the control center unit;

[0023] The airspeed extraction unit, the barometric altitude extraction unit, the metric altitude extraction unit, and the Mach number extraction unit respectively read the green channel pixel values of the airspeed, barometric altitude, metric altitude, and Mach number areas in the video frame, and complete the binarization operation, store the binarization result into the dual-port memory unit, and at the same time send the storage completion flag signal to the control center unit;

[0024] The artificial horizon recognition unit recognizes the white artificial horizon in the video frame, detects the uppermost and lowermost white pixel points of the artificial horizon, and sends the coordinates of these two white pixel points to the roll angle calculation unit;

[0025] The pitch ladder feature point recognition unit detects 4 blue feature points around the pitch ladder symbol in the video frame, and sends the coordinates of the 4 feature points to the rotation calculation unit;

[0026] The roll angle calculation unit calculates the roll angle through the coordinates of the two white pixel points sent by the artificial horizon recognition unit, and sends the roll angle to the result summary unit;

[0027] The rotation calculation unit determines the coordinate area of the pitch ladder through the coordinates of 4 blue feature points sent by the pitch ladder feature point recognition unit, performs bilinear interpolation operation on the pitch ladder through the roll angle sent by the roll angle calculation unit, rotates the pitch ladder to the horizontal state, stores the rotated binary pixel values in the dual-port memory unit, and sends the storage completion flag signal to the control center unit;

[0028] The control center unit receives the storage completion flag signals sent by the airspeed intercept unit, barometric altitude intercept unit, metric altitude intercept unit, Mach number intercept unit and rotation calculation unit, and determines the key symbols to be recognized in the current picture through the timestamp sent by the timestamp recognition unit, and sends the matching start command and the recognized symbol type to the sliding template matching unit;

[0029] After receiving the matching start command and the recognized symbol type sent by the control center unit, the sliding template matching unit reads the binary result of the symbol from the dual-port memory unit, reads the corresponding template of the symbol from the symbol template unit, starts the sliding template matching operation, sends the matching end signal to the control center unit, and sends the sliding template matching result to the result summary unit;

[0030] After receiving the matching end signal of the sliding template matching unit, the control center unit sends a new matching start command and the recognized symbol type to the sliding template matching unit until all the symbols to be recognized in the current picture are completely matched;

[0031] The result summary unit receives the matching results sent by the sliding template matching unit, and outputs the matching results of all symbols after the matching operations of all the key symbols to be monitored in the current picture are completed;

[0032] The binary results of the airspeed, barometric altitude, metric altitude, Mach number and pitch ladder intercept area stored in the dual-port memory unit.

[0033] Beneficial effects

[0034] A time-division multiplexing and sliding template matching head-up display symbol recognition method and system proposed by the present invention, including a timestamp recognition unit, an airspeed interception unit, an atmospheric pressure altitude interception unit, a metric altitude interception unit, a Mach number interception unit, an artificial horizon recognition unit, a roll angle calculation unit, a pitch ladder feature point recognition unit, a rotation calculation unit, a dual-port memory unit, a control center unit, a sliding template matching unit, a symbol template unit, and a result summary unit, is used to solve the recognition problems of six key symbols, namely airspeed, atmospheric pressure altitude, metric altitude, Mach number, roll angle, and pitch ladder, in the head-up display video image. On the one hand, the invention introduced in this method improves the integrity and real-time performance of the symbol recognition method through time-division multiplexing technology, saves computing resources, and reduces hardware costs. On the other hand, the sliding template matching technology is used to improve the accuracy and precision of symbol recognition.

[0035] A head-up display symbol recognition method based on time-division multiplexing and sliding template matching technology proposed by the present invention can accurately recognize key symbols in the head-up display video image by using the method introduced in the present invention, covering a total of six key symbols, namely airspeed, atmospheric pressure altitude, metric altitude, Mach number, roll angle, and pitch ladder, and has the advantages of high integrity, high real-time performance, high precision, low false alarm rate, and low cost. The beneficial effects are as follows:

[0036] (1) By using time-division multiplexing technology, the type of symbol to be recognized in the current video image is determined according to the parity of the timestamp, ensuring that all key symbol recognition tasks are completed within two video frame times, and improving the integrity and real-time performance of symbol recognition operations.

[0037] (2) By using time-division multiplexing technology, all key symbols share the same control center unit and sliding template matching unit to complete symbol recognition operations, greatly reducing the hardware resources required for computing and lowering the hardware cost.

[0038] (3) By using sliding template matching technology, template matching operations are performed multiple times during the process of the symbol interception area sliding on the symbol template to obtain an optimal matching result, greatly improving the accuracy of symbol recognition and reducing the false alarm rate. Description of the Drawings

[0039] Figure 1 is the system composition diagram of the present invention;

[0040] Figure 2 is the schematic diagram of timestamp recognition of the present invention;

[0041] Figure 3 is the schematic diagram of airspeed area interception in the head-up display image of the present invention;

[0042] Figure 4 is the schematic diagram of atmospheric pressure altitude area interception in the head-up display image of the present invention;

[0043] Figure 5 It is a schematic diagram of intercepting the metric height area of the HUD screen described in the present invention;

[0044] Figure 6 It is a schematic diagram of intercepting the Mach number area of the HUD screen described in the present invention;

[0045] Figure 7 It is a schematic diagram of identifying the roll angle of the HUD screen described in the present invention;

[0046] Figure 8 It is a schematic diagram of the position of the characteristic points of the pitch ladder described in the present invention;

[0047] Figure 9 It is a schematic diagram of intercepting the pitch ladder area of the HUD screen described in the present invention;

[0048] Figure 10 It is a schematic diagram of the rotation of the pitch ladder described in the present invention;

[0049] Figure 11 It is a schematic diagram of the sliding template matching process described in the present invention. Detailed implementation manners

[0050] Now, the present invention will be further described in combination with embodiments and drawings:

[0051] A head-up display symbol recognition system based on time-division multiplexing and sliding template matching technology. The head-up display symbol recognition method includes a timestamp recognition unit, an airspeed interception unit, a barometric altitude interception unit, a metric altitude interception unit, a Mach number interception unit, an artificial horizon recognition unit, a roll angle calculation unit, a pitch ladder characteristic point recognition unit, a rotation calculation unit, a dual-port memory unit, a control center unit, a sliding template matching unit, a symbol template unit, and a result summary unit.

[0052] The timestamp recognition unit recognizes the timestamp in the video screen, judges the parity of the timestamp, and sends the timestamp recognition result to the control center unit.

[0053] The airspeed interception unit, the barometric altitude interception unit, the metric altitude interception unit, and the Mach number interception unit respectively read the green channel pixel values in the airspeed, barometric altitude, metric altitude, and Mach number areas of the video screen, and complete the binarization operation, and store the binarization result in the dual-port memory unit.

[0054] After the airspeed interception unit, the barometric altitude interception unit, the metric altitude interception unit, and the Mach number interception unit complete storing the binarization result, they respectively send the storage completion flag signal to the control center unit.

[0055] The artificial horizon recognition unit recognizes the white artificial horizon in the video frame, detects the white pixel points at the topmost and bottommost of the artificial horizon, and sends the coordinates of these two white pixel points to the roll angle calculation unit.

[0056] The roll angle calculation unit calculates the roll angle based on the coordinates of the two white pixel points sent by the artificial horizon recognition unit, and sends the roll angle to the result summary unit.

[0057] The pitch ladder feature point recognition unit detects 4 blue feature points around the pitch ladder symbol in the video frame, and sends the coordinates of the 4 feature points to the rotation calculation unit.

[0058] The rotation calculation unit determines the coordinate area of the pitch ladder through the coordinates of the 4 blue feature points sent by the pitch ladder feature point recognition unit, performs bilinear interpolation operation on the pitch ladder by the roll angle sent by the roll angle calculation unit, rotates the pitch ladder to the horizontal state, and stores the rotated binary pixel values into the dual-port memory unit.

[0059] After storing the rotated binary pixel values, the rotation calculation unit sends a storage completion flag signal to the control center unit.

[0060] The dual-port memory unit stores the airspeed, barometric altitude, metric altitude, Mach number, and the binary result of the pitch ladder intercepted area.

[0061] The control center unit receives the storage completion flag signals sent by the airspeed interception unit, barometric altitude interception unit, metric altitude interception unit, Mach number interception unit, and rotation calculation unit, determines the key symbols to be recognized in the current frame through the timestamp sent by the timestamp recognition unit, and sends a matching start command and the recognized symbol type to the sliding template matching unit.

[0062] After receiving the matching start command and the recognized symbol type sent by the control center unit, the sliding template matching unit reads the binary result of the symbol from the dual-port memory unit, reads the corresponding template of the symbol from the symbol template unit, starts the sliding template matching operation, sends a matching end signal to the control center unit, and sends the sliding template matching result to the result summary unit.

[0063] After receiving the matching end signal from the sliding template matching unit, the control center unit sends a new matching start command and the recognized symbol type to the sliding template matching unit until all the symbols to be recognized in the current frame are completely matched.

[0064] The result summary unit receives the matching results sent by the sliding template matching unit, and outputs the matching results of all symbols after the matching operations of all the key symbols to be monitored in the current frame are completed.

[0065] During recognition, the system receives the video image of the head-up display, determines the parity of the timestamp in the video image, determines the type of key symbol to be recognized in the current image based on the parity of the timestamp, and sequentially completes the recognition operation of the key symbol through the sliding template matching algorithm.

[0066] When the timestamp of the video image is odd, this frame first performs a sliding template matching operation on the airspeed intercept area, and then performs a sliding template matching operation on the Mach number intercept area.

[0067] When the timestamp of the video image is even, this frame first performs a sliding template matching operation on the metric height intercept area, then performs a sliding template matching operation on the pressure altitude intercept area, and finally performs a sliding template matching operation on the pitch ladder intercept area.

[0068] In the sliding template matching algorithm, first, the key symbol area intercepted from the video image is aligned with the upper left corner of the symbol template for template matching operation; after the operation is completed, the intercepted area is moved one column to the right for template matching operation; after the operation is completed, the intercepted area is continuously moved one column to the right until the rightmost column of the intercepted area coincides with the rightmost side of the symbol template for template matching operation; after the operation is completed, the intercepted area coincides with the left side of the symbol template and is moved down one row for template matching operation until the intercepted area coincides with the lower right corner of the symbol template, and the sliding template matching operation ends.

[0069] During the process of the video intercept area moving from the upper left corner to the lower right corner of the symbol template, count the template matching results for each time, find the maximum value of the template matching results and the template area corresponding to the maximum value. The template number in this area is the sliding template matching result of the intercepted area.

[0070] The symbol template needs to cover all situations where the key symbol appears in the video image to ensure that a substantially overlapping area can be found in the corresponding symbol template for the key symbol intercept area of any frame.

[0071] Reference Figure 2 , collect the image information of the head-up display. The first point of the timestamp is at the 2nd - 4th pixel positions in the first row of the image; the second point of the timestamp is at the 22nd - 24th pixel positions; the third point of the timestamp is at the 42nd - 44th pixel positions. Determine whether the gray value of the green channel in the three areas is greater than 100 to complete the recognition and extraction of the image timestamp and determine the parity of the timestamp;

[0072] Reference Figure 3 , obtain the airspeed area data. By determining the left vertex and the right bottom point of the intercept area, sequentially intercept the airspeed data in the units area, tens area, and hundreds area of the airspeed, and store the intercepted image information in the dual-port storage unit;

[0073] Reference Figure 4 For the acquisition of barometric altitude region data, by determining the left vertex and the right bottom point of the intercepted region, successively intercept the barometric altitude data in the barometric altitude tens region, hundreds region, thousands region, and ten thousands region, and store the intercepted image information in the dual-port storage unit;

[0074] Reference Figure 5 For the acquisition of metric altitude region data, by determining the left vertex and the right bottom point of the intercepted region, successively intercept the metric altitude data in the metric altitude tens region, hundreds region, thousands region, and ten thousands region, and store the intercepted image information in the dual-port storage unit;

[0075] Reference Figure 6 For the acquisition of Mach number region data, by determining the left vertex and the right bottom point of the intercepted region, successively intercept the Mach number data in the Mach number tenths region, hundredths region, and thousandths region, and store the intercepted image information in the dual-port storage unit;

[0076] The acquisition of pitch ladder region data mainly has three steps:

[0077] 1) For the calculation of the roll angle, mainly by detecting the white pixel points at the topmost and bottommost of the artificial horizon, obtaining the coordinates of these two white pixel points, and sending the coordinates of these two white pixel points to the roll angle calculation unit, and calculating the roll angle through the arctangent function. This part can refer to Figure 7 ;

[0078] 2) Obtain the coordinates of four feature points around the pitch ladder, as well as the coordinates of the reference point of the intercepted region, refer to Figure 8 . Extract the lower left corner feature point among the four feature points, use this feature point as the rotation center for rotation calculation, and at the same time use the detected topmost feature point as the reference point of the intercepted region. With the abscissa of this reference point as the center, expand 35 pixels to the left and right respectively, and use the ordinate of the reference point as the upper boundary of the intercepted region, and expand 42 pixels downward. The obtained 42×70 region is used as the pitch ladder intercepted region and stored, refer to Figure 9 ;

[0079] 3) The rotation calculation recognition unit performs bilinear interpolation calculation on the pitch ladder intercepted region to obtain the rotated pitch ladder region, refer to Figure 10 , and store the rotated region in the dual-port storage unit.

[0080] Such as Figure 11 , in order to identify the symbol data information of the intercepted region, use the sliding template matching algorithm for matching and recognition. The matching process is as follows:

[0081] 1) Coincide the key symbol area intercepted from the video frame with the upper left corner of the symbol template to complete the template matching operation;

[0082] 2) Move the intercepted area one column to the right to complete the template matching operation;

[0083] 3) Continue to move the intercepted area one column to the right until the rightmost column of the intercepted area coincides with the rightmost side of the symbol template to complete the template matching operation;

[0084] 4) Coincide the intercepted area with the left side of the symbol template and move it down one row to complete the template matching operation;

[0085] 5) Until the intercepted area coincides with the lower right corner of the symbol template, the sliding template matching operation ends;

[0086] 6) During the process of the video intercepted area moving from the upper left corner to the lower right corner of the symbol template, count the template matching results each time, find the maximum value of the template matching results and the template area corresponding to the maximum value. The template number in this area is the sliding template matching result of the intercepted area.

[0087] In this solution, there are various types of matching symbols. In order to control the matching order between symbols, it is necessary to control the central unit for unified allocation. Refer to Figure 11 . Determine the matching order of symbols according to the parity characteristics of the timestamp. When the timestamp is odd, match the airspeed and Mach number. When the timestamp is even, match the metric altitude, pressure altitude, and pitch ladder. The working process of the central control unit is as follows: when the timestamp is odd, after the central control unit receives the airspeed storage completion flag, it sends an airspeed matching start signal to the sliding template matching unit to start the matching operation of the airspeed symbol. After the airspeed matching is completed, the sliding template matching unit sends a matching end signal to the central control unit. After the central control unit receives the Mach number storage completion flag, it sends a Mach number matching start signal to the sliding template matching unit. After receiving the matching end signal from the sliding template matching unit, the central control unit returns to the initial state and waits for the storage completion flag signal of the next frame of the picture. Similarly, when the timestamp is even, the central control unit controls the sliding template matching unit to complete the matching operations of the metric altitude, pressure altitude, and pitch ladder in sequence, and finally obtains the matching results of all symbols.

Claims

1. A time-division multiplexing and sliding template matching method for symbol recognition in a head-up display, characterized in that the steps are as follows: Step 1: Receive the head-up display video image, respectively identify the timestamps in the video image, and at the same time read the green channel pixel values in the airspeed, barometric altitude, metric altitude, and Mach number regions of the video image, and complete the binarization operation to obtain the binarized images of the intercepted regions of the airspeed, barometric altitude, metric altitude, and Mach number regions; At the same time, identify the white artificial horizon in the video image, detect the white pixel points at the topmost and bottommost of the artificial horizon, and calculate the roll angle based on the coordinates of these two white pixel points; At the same time, identify the pitch ladder feature points in the video image, detect the four blue feature points around the pitch ladder symbol in the video image, and obtain the coordinates of the four blue feature points; Step 2: Determine the pitch ladder coordinate region based on the coordinates of the four blue feature points, perform bilinear interpolation operation on the pitch ladder with the roll angle, rotate the pitch ladder to the horizontal state, and obtain the rotated binarized pixel values; Step 3: According to the identified timestamp, judge the parity of the timestamp; When the timestamp is odd, first perform sliding template matching operation on the binarized image of the airspeed intercepted region, and then perform sliding template matching operation on the binarized image of the Mach number intercepted region; When the timestamp is even, first perform sliding template matching operation on the binarized image of the metric altitude intercepted region, then perform sliding template matching operation on the binarized image of the barometric altitude intercepted region, and finally perform sliding template matching operation on the pitch ladder intercepted region; The sliding template matching method: First, overlap the key symbol region intercepted from the video image with the upper left corner of the symbol template for template matching operation; after the operation is completed, move the intercepted region one column to the right for template matching operation; after the operation is completed, continue to move the intercepted region one column to the right until the rightmost column of the intercepted region coincides with the rightmost side of the symbol template for template matching operation; after the operation is completed, overlap the intercepted region with the left side of the symbol template and move down one row for template matching operation until the intercepted region coincides with the lower right corner of the symbol template, and the sliding template matching operation ends; Step 4: When the sliding template matching operation ends, all the symbols to be recognized in the current image are matched, and the matching results of all symbols are obtained; When the next frame of image needs to be matched, repeat Steps 1 to 3.

2. The time-division multiplexing and sliding template matching method for symbol recognition in a head-up display according to claim 1, characterized in that: During the process of the video intercepted region moving from the upper left corner to the lower right corner of the symbol template, for each template matching result, find the maximum value of the template matching result and the template region corresponding to the maximum value, and the template number in this region is the sliding template matching result of the intercepted region.

3. The time-division multiplexing and sliding template matching method for symbol recognition in a head-up display according to claim 2, characterized in that: The symbol template needs to cover all situations where the key symbols appear in the video image, ensuring that the key symbol intercepted regions of any frame can find a substantially overlapping region in the corresponding symbol template.

4. A time-division multiplexing and sliding template matching head-up display symbol recognition method according to claim 1, characterized in that: Obtaining the roll angle: Detect the white pixel points at the uppermost and lowermost positions of the artificial horizon, and calculate the roll angle through the arctangent function.

5. A time-division multiplexing and sliding template matching head-up display symbol recognition method according to claim 1, characterized in that: The pitch ladder intercepting area: Extract the lower left corner feature point among the four feature points around the pitch ladder, use this feature point as the rotation center for rotation calculation, and at the same time use the detected uppermost feature point as the reference point for the intercepting area. With the abscissa of this reference point as the center, expand 35 pixels to the left and right respectively, and use the ordinate of the reference point as the upper boundary of the intercepting area, and expand 42 pixels downward. The obtained 42×70 area is the pitch ladder intercepting area.

6. A system for implementing the time-division multiplexing and sliding template matching head-up display symbol recognition method according to any one of claims 1 to 5, characterized in that it includes a timestamp recognition unit, an airspeed intercepting unit, an atmospheric pressure altitude intercepting unit, a metric altitude intercepting unit, a Mach number intercepting unit, an artificial horizon recognition unit, a roll angle calculation unit, a pitch ladder feature point recognition unit, a rotation calculation recognition unit, a dual-port memory unit, a control center unit, a sliding template matching unit, a symbol template unit and a result summary unit; the timestamp recognition unit, the airspeed intercepting unit, the atmospheric pressure altitude intercepting unit, the metric altitude intercepting unit, the Mach number intercepting unit and the artificial horizon recognition unit simultaneously receive the input image information, where the timestamp recognition unit recognizes the timestamp in the video frame, judges the parity of the timestamp, and sends the timestamp recognition result to the control center unit; the airspeed intercepting unit, the atmospheric pressure altitude intercepting unit, the metric altitude intercepting unit and the Mach number intercepting unit respectively read the green channel pixel values of the airspeed, atmospheric pressure altitude, metric altitude and Mach number areas in the video frame, and complete the binarization operation, store the binarization result in the dual-port memory unit, and at the same time send the storage completion flag signal to the control center unit; the artificial horizon recognition unit recognizes the white artificial horizon in the video frame, detects the white pixel points at the uppermost and lowermost positions of the artificial horizon, and sends the coordinates of these two white pixel points to the roll angle calculation unit; the pitch ladder feature point recognition unit detects 4 blue feature points around the pitch ladder symbol in the video frame, and sends the coordinates of the 4 feature points to the rotation calculation unit; the roll angle calculation unit calculates the roll angle through the coordinates of the two white pixel points sent by the artificial horizon recognition unit, and sends the roll angle to the result summary unit; the rotation calculation unit determines the pitch ladder coordinate area through the coordinates of the 4 blue feature points sent by the pitch ladder feature point recognition unit, performs bilinear interpolation operation on the pitch ladder through the roll angle sent by the roll angle calculation unit, rotates the pitch ladder to the horizontal state, stores the rotated binarized pixel values in the dual-port memory unit, and sends the storage completion flag signal to the control center unit; The control center unit receives the storage completion flag signals sent by the airspeed intercept unit, the barometric altitude intercept unit, the metric altitude intercept unit, the Mach number intercept unit, and the rotation calculation unit, and determines the key symbols to be recognized in the current picture through the timestamps sent by the timestamp recognition unit, and sends the matching start command and the recognized symbol type to the sliding template matching unit; After receiving the matching start command and the recognized symbol type sent by the control center unit, the sliding template matching unit reads the binarization result of the symbol from the dual-port memory unit, reads the corresponding template of the symbol from the symbol template unit, starts the sliding template matching operation, sends the matching end signal to the control center unit, and sends the sliding template matching result to the result summary unit; After receiving the matching end signal of the sliding template matching unit, the control center unit sends a new matching start command and the recognized symbol type to the sliding template matching unit until all the symbols to be recognized in the current picture are completely matched; The result summary unit receives the matching results sent by the sliding template matching unit, and outputs the matching results of all symbols after the matching operations of all the key symbols to be monitored in the current picture are completed; The binarization results of the airspeed, barometric altitude, metric altitude, Mach number, and pitch ladder intercept areas stored in the dual-port memory unit.

Images