Real-time image stabilization method and system for intelligent image processing

Abstract

The invention provides a real-time image stabilization method and system for intelligent image processing, and relates to the technical field of image stabilization. The method comprises the following steps: synchronously obtaining continuous image frames under strong glare interference, and identifying a texture blurred region; for the region, calculating a texture complexity distribution diagram by combining the local structure variation of the adjacent frames, and screening candidate regions according to the texture complexity distribution diagram; polarization filtering processing is applied to the candidate area, the filtering angle is dynamically adjusted to be matched with the polarization direction of incident light, and a polarization filtering image is generated; extracting a candidate region motion vector field based on the gradient continuity feature of the image, and performing spatial superposition on the candidate region motion vector field and the texture complexity distribution map to form a motion vector set carrying texture weight factors; and according to the texture weight factor, carrying out hierarchical clustering to separate a background vector group and a foreground abnormal group, only selecting a geometric center value of the background group to execute inter-frame motion compensation, and outputting a stable image sequence. According to the invention, the stability and quality of images collected by the unmanned intelligent turntable in a strong glare dynamic scene are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of image stabilization technology, in particular to a real-time image stabilization method and system for intelligent image processing. BACKGROUND

[0002] In a complex light environment, the imaging device mounted on an unmanned intelligent turntable often produces high-frequency image jitter due to severe pitching and rotating motion and environmental strong glare interference, resulting in inaccurate motion target tracking and distorted scene recognition. Such a scene urgently needs a real-time image stabilization technology that can overcome both the texture feature degradation caused by dynamic glare and the inter-frame displacement distortion caused by turntable motion, while maintaining low processing delay to meet real-time control requirements.

[0003] The current mainstream solution uses multi-scale optical flow field fusion and adaptive motion filtering technology: by extracting the dense optical flow field of adjacent frames, the motion vector level is divided in combination with scene depth information; then a space-time filter is constructed based on the statistical distribution of motion vector amplitude to suppress high-frequency abnormal motion components; finally, inter-frame compensation is achieved through global motion model fitting. This solution relies on GPU acceleration to ensure the real-time performance of optical flow calculation.

[0004] This solution has double failures in strong glare environment: glare overwhelms local texture structure, resulting in a large number of erroneous estimation points in the texture sparse area of the optical flow field, and the motion vector level division is inaccurate; the transient light intensity changes caused by environmental glare are misidentified as foreground motion noise, and the space-time filter incorrectly filters out the real turntable motion components, causing distortion of the compensation model. This defect is particularly pronounced in high-reflectivity scenes, and the stabilized image still has periodic jitter and edge tearing. SUMMARY

[0005] The present application provides a real-time image stabilization method and system for intelligent image processing to solve the problem of low stabilization quality of images collected by an unmanned intelligent turntable in the prior art.

[0006] In a first aspect, the present application provides a real-time image stabilization method for intelligent image processing, comprising: When the unmanned intelligent turntable performs pitching or rotating motion, continuously acquiring image frames disturbed by strong environmental glare, and identifying texture blur areas caused by light reflection in the image frames; For the texture blur area, calculating a texture complexity distribution map in combination with the local structure change amount of adjacent image frames, and selecting a candidate area with a texture feature retention rate higher than a preset standard in the texture blur area based on the texture complexity distribution map; applying a polarized light filtering process to the candidate region, generating a polarized light filtering image by adjusting a filtering angle to match an incident light polarization direction of a current frame, the polarized light filtering image simultaneously having a gradient continuity feature of suppressing glare noise and enhancing edge gradient; based on the gradient continuity feature of the polarized light filtering image, extracting a motion vector field of the candidate region between adjacent frames, and spatially superimposing the motion vector field and the texture complexity distribution map to form a motion vector set carrying a texture weight factor; According to the texture weight factor, the motion vector set is hierarchically clustered to separate the background motion vector group generated by the motion of the unmanned intelligent turntable itself and the foreground abnormal vector group caused by environmental glare interference, and only the geometric center value of the background motion vector group is selected for inter-frame motion compensation to output a stable image sequence.

[0007] Optionally, based on the gradient continuity feature of the polarized light filtering image, the motion vector field of the candidate region between adjacent frames is extracted, and the motion vector field is spatially superimposed with the texture complexity distribution map to form a motion vector set carrying a texture weight factor, comprising: In the candidate region of the polarized light filtering image, the displacement trajectory of the corresponding edge points between adjacent frames is tracked along the object edge direction, and a motion vector field is formed by the direction and length of the displacement trajectory; The spatial position of each motion vector in the motion vector field is bound with the value of the texture complexity distribution map at the same spatial position; According to the bound value, each motion vector is assigned a texture weight factor, and the texture weight factor is in a proportional relationship with the value of the texture complexity distribution map, forming a motion vector set carrying a texture weight factor.

[0008] Optionally, the polarized light filtering process is applied to the candidate region, and a polarized light filtering image is generated by adjusting the filtering angle to match the incident light polarization direction of the current frame, the polarized light filtering image simultaneously having a gradient continuity feature of suppressing glare noise and enhancing edge gradient, comprising: According to the light reflection characteristics of the candidate region, the dominant polarization direction of the incident light in the candidate region is analyzed, and the dominant polarization direction is determined by the periodic change mode of the pixel brightness in consecutive frames; Dynamically rotating the light transmission axis angle of the physical filter to form an orthogonal relationship between the light transmission axis angle and the dominant polarization direction; According to the adjusted light transmission axis angle, the light of the candidate region passing through the physical filter is collected to generate a polarized light filtering image, and the brightness amplitude of the glare noise in the polarized light filtering image is reduced and the light and dark transition continuity of the object edge is improved.

[0009] Optionally, the motion vector set is hierarchically clustered according to the texture weight factor, separating a background motion vector group caused by the motion of the unmanned intelligent turntable itself and a foreground abnormal vector group caused by environmental glare interference, only selecting the geometric center value of the background motion vector group for inter-frame motion compensation, and outputting a stable image sequence, comprising: The motion vector set is divided into a high weight group and a low weight group according to the texture weight factor, the high weight group containing motion vectors with texture weight factors higher than a preset separation threshold, and the low weight group containing the remaining motion vectors; The direction consistency density of the motion vectors in the high weight group is calculated, and the motion vectors falling into the density peak value region are classified as a background motion vector group, and the low weight group and the motion vectors in the high weight group not falling into the density peak value region are classified as a foreground abnormal vector group; Only the geometric center value of all vectors in the background motion vector group is selected as the translation reference, and reverse displacement compensation is performed on the next frame of image to output a stable image sequence.

[0010] Optionally, for the texture blur area, a texture complexity distribution map is calculated in combination with the local structure change amount of the adjacent image frames, and a candidate area with a texture feature retention rate higher than a preset standard in the texture blur area is selected based on the texture complexity distribution map, comprising: For each image unit in the texture blur area, the structure change amount between the corresponding image units of the current frame and the adjacent frame is calculated, and the structure change amount is represented by the cumulative value of the difference amplitude of the pixel values within the unit; The structure change amount is mapped to each image unit position of the texture blur area to generate a texture complexity distribution map, and the value of the texture complexity distribution map reflects the strength of the ability of the image unit to resist glare interference; In the texture complexity distribution map, image units with values higher than a preset threshold are marked, and continuously distributed marked units in the image units are merged into a candidate area, and the candidate area covers a continuous spatial range with a texture feature retention rate higher than a preset standard.

[0011] Optionally, the light transmission axis angle of the dynamic rotating physical filter is rotated to form an orthogonal relationship with the dominant polarization direction, comprising: The angle value of the dominant polarization direction is measured, and the orthogonal angle gap between the angle value and the light transmission axis angle is calculated; The number of motor driving pulses is generated according to the orthogonal angle gap, and the rotating motor is driven to rotate the physical filter by a gap compensation angle; During rotation, when the light transmission axis angle sensor feedback value reaches the orthogonal angle of the angle value, the rotating motor is stopped and the light transmission axis is locked.

[0012] Optionally, the texture weight factor is assigned to each motion vector according to the bound value, the texture weight factor is in direct proportion to the value of the texture complexity distribution map, and a motion vector set carrying the texture weight factor is formed, including: The bound value and the value of the texture complexity distribution map corresponding to the spatial position of the motion vector are taken as the texture reliability value of the current motion vector; The texture reliability value is input into a linear converter to generate a texture weight factor through a fixed magnification operation, and the value of the texture weight factor is equal to the texture reliability value multiplied by a preset proportion coefficient; The texture weight factor is written into the attribute field of the corresponding motion vector to construct a motion vector set carrying the texture weight factor.

[0013] In a second aspect, the present application provides a real-time image stabilization system for intelligent image processing, comprising: An acquisition module is configured to acquire continuous image frames disturbed by environmental glare when an unmanned intelligent turntable performs pitching or rotating motion, and identify texture blur areas caused by light reflection in the continuous image frames; A calculation module is configured to calculate a texture complexity distribution map in combination with the local structure change amount of adjacent image frames for the texture blur areas, and filter candidate areas with a texture feature retention rate higher than a preset standard in the texture blur areas based on the texture complexity distribution map; A processing module is configured to apply polarized light filtering processing to the candidate areas, adjust the filtering angle to match the incident light polarization direction of the current frame, and generate a polarized light filtering image, which has the characteristics of suppressing glare noise and enhancing gradient continuity of edge gradient; An extraction module is configured to extract a motion vector field of the candidate areas between adjacent frames based on the gradient continuity feature of the polarized light filtering image, and perform spatial superposition of the motion vector field and the texture complexity distribution map to form a motion vector set carrying a texture weight factor; A separation module is configured to hierarchically cluster the motion vector set according to the texture weight factor, separate a background motion vector group generated by the motion of the unmanned intelligent turntable itself and a foreground abnormal vector group caused by environmental glare disturbance, and only select the geometric center value of the background motion vector group for inter-frame motion compensation to output a stable image sequence.

[0014] In a third aspect, the present application provides a computing device, comprising a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement the real-time image stabilization method for intelligent image processing according to the first aspect.

[0015] In a fourth aspect, the present application provides a computer storage medium, which stores a computer program; when the computer program is executed by a computer, a real-time image stabilization method for intelligent image processing according to the first aspect is implemented.

[0016] In the embodiments of the present application, when the unmanned intelligent turntable performs pitching or rotating motion, continuous image frames disturbed by strong environmental glare are synchronously acquired, and a texture blur area caused by light reflection in the continuous image frames is identified; for the texture blur area, a texture complexity distribution map is calculated in combination with a local structure change amount of adjacent image frames, and a candidate area with a texture feature retention rate higher than a preset standard is screened based on the texture complexity distribution map; a polarized light filtering process is applied to the candidate area, a polarized light filtering image is generated by adjusting a filtering angle to match an incident light polarization direction of a current frame, the polarized light filtering image has both glare noise suppression and gradient continuity features of enhanced edge gradient; based on the gradient continuity features of the polarized light filtering image, a motion vector field of the candidate area between adjacent frames is extracted, and the motion vector field is spatially superimposed with the texture complexity distribution map to form a motion vector set carrying a texture weight factor; the motion vector set is hierarchically clustered according to the texture weight factor, a background motion vector group generated by the motion of the unmanned intelligent turntable itself and a foreground abnormal vector group caused by environmental glare disturbance are separated, only the geometric center value of the background motion vector group is selected for inter-frame motion compensation, and a stable image sequence is output.

[0017] The technical scheme of the present application has the following beneficial effects: The texture degradation area in the image under strong glare interference is accurately positioned, which provides a targeted target for subsequent processing and avoids waste of global computing resources. By quantifying and screening the texture complexity, a candidate area with strong anti-glare ability is extracted to ensure the reliability of the motion estimation basic features. Dynamic polarization filtering generates an optimized image with glare suppression and edge enhancement, directly solving the texture blur problem and providing high continuity features for motion tracking. The texture reliability and motion vector are fused to form a weighted motion set, and an anti-interference motion analysis input is constructed. Based on the texture weight-based clustering, the real motion of the turntable and the glare pseudo-motion are accurately separated to ensure the physical consistency of the motion compensation.

[0018] Further, in the candidate region of the polarization filtered image, a motion vector field is constructed by tracking the displacement trajectory of adjacent frame edge points along the object edge; the spatial position of each motion vector is bound with the texture complexity distribution map value at the same position; a texture weight factor is assigned to each motion vector in a proportional relationship according to the bound value, and a set of weighted motion vectors is formed. Through the spatial binding and weight mapping of motion vectors and texture complexity, the texture anti-interference ability is converted into a motion vector credibility index, which provides a decision basis for distinguishing real motion from glare false motion for subsequent clustering.

[0019] These aspects or other aspects of the present application will be made clearer in the following description of embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor under the premise of the drawings.

[0021] Figure 1 A flow chart of a real-time image stabilization method for intelligent image processing provided by the present application is shown; Figure 2 A flow chart of a real-time image stabilization method for intelligent image processing provided by the present application is shown; Figure 3 A structural schematic diagram of a real-time image stabilization system for intelligent image processing provided by the present application is shown; Figure 4 A structural schematic diagram of a computing device provided by the present application is shown. DETAILED DESCRIPTION

[0022] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application.

[0023] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this document or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any order of execution. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this document are used to distinguish different messages, devices, modules, etc., and do not represent a sequential order, nor do they limit "first" and "second" to being different types.

[0024] Current image stabilization solutions based on multi-scale optical flow fusion and adaptive motion filtering face fundamental limitations when unmanned intelligent turntables encounter strong glare. First, glare overwhelms local texture structures, causing the optical flow field to generate a large number of erroneous motion estimation points in sparsely textured areas, undermining the physical authenticity of the motion vector hierarchy. Second, transient light intensity changes caused by ambient glare are misidentified as foreground motion noise, causing the spatiotemporal filter to incorrectly filter out the turntable's true motion components. This dual flaw causes the compensation model to output distorted images with periodic jitter and edge tearing in highly reflective scenes, severely limiting the reliability of real-time operation in complex lighting environments.

[0025] In response to the above-mentioned defects, the present invention proposes a real-time stabilization method for texture-polarization-motion collaboration, the core of which is to reconstruct the reliability basis of motion estimation through cross-modal coupling of dynamic polarization filtering and texture anti-interference ability. Specifically, under strong glare interference, candidate areas with high texture feature retention rates are first screened and polarization angle adaptive filtering is applied to generate an optimized image with continuous edge gradients; then, the texture complexity is converted into a motion vector weight factor to drive hierarchical clustering to accurately separate the real motion of the turntable from the glare pseudo-motion; finally, inter-frame compensation is performed based on the geometric center value of the background motion group. This method has made a breakthrough in solving the problem of motion estimation distortion caused by texture feature degradation in glare scenes - it not only eliminates the erroneous estimation of the optical flow field in texture sparse areas, but also avoids the mis-filtering of motion components caused by transient light changes, fundamentally overcoming the periodic jitter and edge tearing defects in the prior art, while meeting the millisecond-level real-time processing requirements.

[0026] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of this application.

[0027] Figure 1A flowchart of a real-time image stabilization method for intelligent image processing is provided for embodiments of the present application, as shown in Figure 1 The method comprises: 101. When the unmanned intelligent turntable performs pitching or rotating motion, continuously acquire image frames disturbed by strong environmental glare, and identify texture blur areas caused by light reflection in the image frames; In the above scheme, the unmanned intelligent turntable is a rotatable platform carrying a camera device, which can perform up-down tilting or horizontal rotating motion. The strong environmental glare disturbance is a dazzling light spot formed by strong natural light or artificial light on the surface of high-reflective objects such as metal and water. The continuous image frames are multiple images continuously captured in time sequence during the turntable motion. The texture blur area is an area in which the details such as the surface texture and edge of an object are lost due to glare coverage in the image.

[0028] In the embodiments of the present application, first, the image acquisition module is triggered by the attitude sensor built in the turntable, and a high-speed camera is started synchronously at the moment when the pitching or rotating motion of the turntable starts, to capture continuous images at a rate of 30 frames per second. If there is a strong glare source in the environment at this time, the camera sensor will capture an image sequence containing overexposed spots.

[0029] Secondly, the acquired continuous image frames are analyzed by region scanning: each image is divided into a plurality of square units, such as 16x16 pixel blocks; the pixel value fluctuation amplitude of each unit is calculated, i.e. the difference between the brightest and darkest pixels in the unit; the unit with a fluctuation amplitude lower than a set threshold is marked as a low-texture unit, and such area loses the contrast between light and dark due to glare coverage, presenting a uniform bright spot; the low-texture units that are continuously adjacent in space are merged into a texture blur area. Finally, the boundary coordinate set of all texture blur areas is output as the targeted target for subsequent processing.

[0030] 102. For the texture blur area, calculate a texture complexity distribution map in combination with the local structure change amount of the adjacent image frames, and filter candidate areas in the texture blur area with a texture feature retention rate higher than a preset standard based on the texture complexity distribution map; Optionally, as shown in Figure 2 The step 102 can specifically include the following steps: 1021. For each image unit in the texture blur area, calculate the structure change amount of the corresponding image unit between the current frame and the adjacent frame, which is represented by the cumulative value of the difference amplitude of the pixel values within the unit; 1022. Map the structure change amount to each image unit position of the texture blur area to generate a texture complexity distribution map, and the value of the texture complexity distribution map reflects the strength of the ability of the image unit to resist glare disturbance; 1023、In the texture complexity distribution map, mark the image unit whose value is higher than the preset threshold, combine the continuously distributed marked units in the image unit into a candidate region, and the candidate region covers a continuous spatial range with a texture feature retention rate higher than a preset standard.

[0031] In the above scheme, the image unit refers to a fixed-size square region divided by the texture blur region. The structure variation amount refers to the sum of the brightness variation of all pixels of the image unit at the same position in two adjacent frames. The larger the value is, the more obvious the texture detail variation is. The texture complexity distribution map is a mapping diagram marking the structure variation amount value of each image unit. The high-value area represents a position with strong anti-glare ability. The candidate region is a continuous block formed by the adjacent connection of high-value image units. These regions can still retain the real texture features of the object under the glare interference.

[0032] In the embodiment of the application, first, the texture blur region identified in step 101 is divided into fixed-size squares through step 1021. The following operations are performed on each square: locating the corresponding square of the adjacent frame: aligning the same physical region of the current frame and the previous frame according to the rotation table motion parameters, pitch / rotation angle; calculating the pixel brightness difference: comparing the brightness values of the corresponding positions in the two frame squares pixel by pixel, and calculating the absolute value difference, for example, the brightness of pixel A changes from 120 to 115, the difference = |120-115|=5; accumulating the difference value: adding the brightness difference values of all pixels in the square to obtain the structure variation amount. The larger the total brightness variation is, the richer the texture details of the region are, and the smaller the glare interference is.

[0033] Then, spatial mapping is performed through step 1022: a coordinate label is generated for each square, and the structure variation amount value is marked at the corresponding coordinate position; a distribution map is generated: the values of all squares are arranged according to the spatial position to form a two-dimensional value matrix; and the ability strength is associated: the region with a value ≥1000 in the distribution map represents strong anti-glare ability and clear texture, and the region with a value ≤500 represents weak anti-glare ability and glare submerged texture.

[0034] Finally, threshold marking is performed through step 1023: set the anti-glare ability threshold, mark all squares with a value ≥1000 in the distribution map; adjacent merging: check the continuity of adjacent marked squares, i.e. up, down, left and right connection, combine the continuous blocks into candidate regions, for example, three adjacent squares with values 1280, 1200 and 1100 are combined into an L-shaped region; region output: the candidate region needs to meet the minimum area constraint to ensure covering the effective texture area.

[0035] In practical application, the vehicle travels on a mountain road at night at 60 km / h, and the vehicle-mounted camera continues to shake due to the uneven road surface. The motorcycle taillight in front forms a candidate region in the monitoring picture. The region is divided into 256 16x16 pixel squares through step 1021. For a square at the position: left edge of the bumper, the calculation is: the average brightness of the current frame pixel is 170, and the last frame is 163; the maximum brightness difference of a single pixel is |182-175|=7, and the minimum difference is |155-152|=3; the total brightness difference of 256 pixels is 1350, which is high because the reflection of the concave indentation of the bumper changes).

[0036] Step 1022 scene marks 1350 on the coordinates (120, 80) of the square, and the adjacent glare center square, position: middle of the bumper, calculates the brightness difference sum of only 280, and the pixel brightness under the glare is almost unchanged. The generated distribution chart shows that the square value of the bumper edge is 1100-1500, which has high anti-glare ability; the square value of the glare center is 200-400, which has low anti-glare ability.

[0037] Step 1023 scene sets the threshold value 1000, and marks the high value square. Merge the 12 squares with continuous distribution on the edge of the bumper, with values of 1100-1500, to form a U-shaped candidate region around the glare spot, which clearly retains the license plate characters and screw hole texture. Anti-vibration verification: in the process of continuous shaking of the vehicle, the same physical region in the adjacent frame is shifted by 5 pixels due to vibration. After correcting the coordinates by the turntable motion parameters, the structure change calculation is not affected by the displacement.

[0038] The overall scheme of the above step 102, when the high-speed moving carrier is faced with the double interference of vibration and glare, accurately locks the physical region with the smallest glare influence by quantifying the dynamic change characteristics of the local texture, and builds an anti-interference anchor point basis for subsequent motion analysis, which fundamentally avoids the feature distortion problem caused by vibration displacement and strong light flooding.

[0039] 103, apply polarization filtering processing to the candidate region, adjust the filtering angle to match the incident light polarization direction of the current frame, generate a polarization filtering image, and the polarization filtering image has the characteristics of suppressing glare noise and enhancing edge gradient gradient continuity; Optionally, step 103 can specifically include the following steps: 1031, according to the light reflection characteristics of the candidate region, analyze the dominant polarization direction of the incident light in the candidate region, and the dominant polarization direction is determined by the periodic change mode of the pixel brightness in the continuous frame; 1032, dynamically rotate the light transmission axis angle of the physical filter, so that the light transmission axis angle forms an orthogonal relationship with the dominant polarization direction; The step 1032 can specifically include the following processes: measuring the angle value of the dominant polarization direction, calculating the orthogonal angle gap between the angle value and the angle of the light transmission axis; generating the number of motor driving pulses according to the orthogonal angle gap, driving the rotary motor to rotate the physical filter by the gap compensation angle; when the feedback value of the light transmission axis angle sensor reaches the orthogonal angle of the angle value during the rotation process, stopping the rotary motor and locking the light transmission axis.

[0040] 1033. According to the adjusted angle of the light transmission axis, the light rays passing through the candidate region of the physical filter are collected to generate a polarized filter image, in which the brightness amplitude of the glare noise is reduced and the continuity of the light and dark transition of the object edge is improved.

[0041] In the above scheme, the dominant polarization direction refers to the main vibration angle of the reflected light in the candidate region. The angle of the light transmission axis is the direction angle of the physical filter allowing light to pass through. The orthogonal relationship refers to the perpendicularity between the light transmission axis and the polarization direction. The brightness amplitude refers to the range of the sharp fluctuation of the pixel brightness of the glare region. The light and dark transition continuity refers to the smoothness of the color depth change at the object edge.

[0042] In the embodiments of the present application, first, the continuous frame brightness sequence is extracted through step 1031, the brightness value of each pixel in the candidate region in the continuous 5 frames of images is selected, and the time dimension brightness sequence is generated; the brightness fluctuation direction is calculated: the sign change rule of the brightness difference value of each pixel is calculated: if the brightness value periodically fluctuates as "low-high-low-high", the fluctuation direction of the pixel is recorded as 0°; if it fluctuates as "high-low-high-low", it is recorded as 90°; other rules are considered as no direction. The dominant direction is counted: all effective directions in the candidate region are classified into the 0°-180° angle interval, and the median angle of the group with the highest proportion is counted as the dominant polarization direction. For example, the 5 frame brightness of a certain pixel [50, 120, 55, 125, 60]→difference sequence [+70, -65, +70, -65]→sign sequence [+,-,+,-]→fluctuation direction 0°.

[0043] Then, the target angle is calculated through step 1032: when the dominant polarization direction is α, the target angle of the light transmission axis = α+90°, and if it exceeds 180°, it is reduced by 180°. Drive the rotating mechanism: read the current angle β of the filter; calculate the shortest rotation path: if |α+90°-β|>180°, rotate in the opposite direction; send pulse instruction: rotate 0.1° per pulse, total pulse number = angle difference / 0.1. Real-time closed-loop calibration: the actual angle is fed back in real time through the angle sensor, and when the error with the target angle is less than 1°, the rotation is stopped. For example, the dominant direction α=40°→target angle=130°; the current angle β=20°→110° needs to be rotated→1100 pulses are sent.

[0044] Finally, step 1033 performs physical light filtering: when the transmission axis is orthogonal to the polarization direction, the main component of glare is blocked. Low-light compensation exposure: Because filtering reduces incoming light, the exposure time is automatically extended by 1.5 times. Image generation verification: In glare areas, the brightness value changes from violent fluctuations to stability; at object edges, the previously discontinuous grayscale transitions become continuous.

[0045] In a practical application, suppose a surveillance camera captures a wet road at night, with headlights forming a candidate region near a pool of water. Step 1031 analyzes five consecutive frames at 0.05-second intervals: pixel A's brightness sequence: [180, 50, 190, 55, 200] → differences: [+130, -140, +135, -145] → signs: [+, -, +, -] → fluctuation direction: 0°. 68% of the pixels in the region fluctuate at 0°, indicating a dominant polarization direction of 0°.

[0046] Through step 1032, the target angle of the transmission axis = 0° + 90° = 90°; the current angle sensor reading is 170°, and the shortest path is: 80° counterclockwise, 170° → 180° → 0° → 90°; 800 pulses are sent, that is, 80° / 0.1°. Vibration during rotation causes an offset of 3°, and 30 pulses are added to compensate for the offset → the final angle is 90.2°.

[0047] After filtering in step 1033, the image capture shows the following: the brightness of the water reflection drops from an average of 200 to 60; the edge of the zebra crossing, previously obscured by headlight glare, becomes visible, with the grayscale value changing from [25, 155] to [65, 68, 71]; and the texture details of the pedestrian's shoe sole become visible, with the brightness difference originally <5 now >15. Low-light compensation: The exposure time is extended from 1 / 100 second to 1 / 67 second, with no significant increase in dark noise.

[0048] The overall solution of step 103 above, in night scenes where weak light and strong glare coexist, accurately suppresses interfering reflections of artificial light sources through dynamic polarization filtering, and simultaneously enhances the visual continuity of object edges in low-light environments, providing a clear image base for moving target tracking that is not contaminated by ambient light, significantly improving the image stabilization reliability of the nighttime monitoring system.

[0049] 104. Extracting a motion vector field of the candidate region between adjacent frames based on a gradient continuity feature of the polarization filtered image, and spatially superimposing the motion vector field with the texture complexity distribution map to form a motion vector set carrying a texture weight factor; Optionally, step 104 may specifically include the following steps: 1041. In the candidate area of ​​the polarization filter image, track the displacement trajectory of the corresponding edge point between adjacent frames along the edge of the object, and construct a motion vector field based on the direction and length of the displacement trajectory; 1042. Bind the spatial position of each motion vector in the motion vector field to the value of the texture complexity distribution map at the same spatial position; 1043. Assign a texture weight factor to each motion vector according to the bound value, wherein the texture weight factor is proportional to the value of the texture complexity distribution map, thereby forming a set of motion vectors carrying texture weight factors.

[0050] Among them, step 1043 may specifically include the following process: using the bound value and the texture complexity distribution map value corresponding to the spatial position of the motion vector as the texture credibility value of the current motion vector; inputting the texture credibility value into the linear converter, and generating a texture weight factor through a fixed-rate amplification operation, the value of the texture weight factor is equal to the texture credibility value multiplied by a preset proportional coefficient; writing the texture weight factor into the attribute field of the corresponding motion vector, and constructing a motion vector set carrying the texture weight factor.

[0051] In the above scheme, the gradient continuity feature is the smooth transition characteristic of color changes at the edge of an object. The displacement trajectory is the movement path of the same object's feature point between two frames. The motion vector field is a collection of multiple directional arrows, each arrow representing the direction and distance of movement of a feature point. Spatial binding establishes a correspondence between the position coordinates of the arrows and the anti-interference ability value of the texture at the same location. The texture weight factor is a numerical measure of the reliability of the motion vector; the stronger the texture's anti-interference ability, the higher the value.

[0052] In this embodiment of the present application, step 1041 is first performed to locate edge points: the color mutation position is scanned within the candidate area of ​​the polarization filter image and recorded as an edge feature point. Cross-frame trajectory matching: in the previous frame image, with the current edge point as the center, the most similar point is searched within a 5×5 pixel range: color features, i.e., RGB value similarity, are compared; shape features, i.e., adjacent pixel gradient patterns, are compared; the point with the highest similarity is selected as the matching point; vector field construction: the coordinate difference (△x, △y) between the current point and the matching point is calculated, and an arrow pointing from the previous frame position to the current frame position is generated: arrow direction = arctan(△y / △x); arrow length = (△x²+△y²)^(1 / 2). For example, the edge point (205,318) in the current frame is matched to the previous frame (202,315), generating an arrow with a direction of 45° and a length of 4.2 pixels.

[0053] By step 1042, read the starting point coordinate data of each motion vector in the motion vector field, that is, the position information of the current frame edge feature point, then retrieve the pre-stored texture complexity value at the corresponding coordinate position of the texture complexity distribution map, establish the binding relationship table of motion vector identifier and texture value, which is stored in text format as "motion vector ID, coordinate X, coordinate Y, texture value" entries, for example, "MV102, 355, 422, 1350" represents the motion vector with identifier MV102 binds the texture value 1350 at coordinate (355, 422).

[0054] By step 1043, based on the texture value in the binding relationship table, calculate the weight factor through the preset linear conversion coefficient, the calculation process: texture value x coefficient, then create a new data structure containing displacement data and weight for each motion vector, generate records in the format of "motion vector ID, horizontal displacement, vertical displacement, weight factor", for example, MV102 binds value 1350 → weight 1.35 → record as "MV102, 3, 3, 1.35", finally aggregate all records in coordinate order to form a comma-separated text collection, such as "MV102, 3, 3, 1.35; MV205, 1, 0, 0.32;...".

[0055] In actual application, in the night highway toll monitoring, through step 1041, for the right lower corner edge point of the license plate character after polarization filtering, the current frame coordinate (355, 422), match to the best point (352, 419) in the 5x5 pixel search area of the last frame, calculate the displacement difference Δx=3, Δy=3 to generate a 45-degree direction / 4.2-pixel length motion vector.

[0056] By step 1042, read the starting point coordinate (355, 422) of the vector, query the texture distribution map to obtain the same position value 1350, and generate the binding record "MV102, 355, 422, 1350".

[0057] By step 1043, calculate the weight 1350x0.001=1.35, construct the vector record "MV102, 3, 3, 1.35", and process the glare area vector at the same time, coordinate (280, 380) texture value 320 → weight 0.32 → record "MV205, 1, 0, 0.32", output the complete collection: "MV102, 3, 3, 1.35; MV205, 1, 0, 0.32;...".

[0058] The overall scheme of the above step 104, in weak light environment, by fusing the polarization-enhanced edge features and texture anti-interference ability, constructs a motion vector set with credibility weight, provides a reliability quantization basis for subsequent motion analysis, and significantly improves the stabilization accuracy of night motion target.

[0059] 105. hierarchically clustering the set of motion vectors according to the texture weight factor, separating a background motion vector group caused by the motion of the unmanned intelligent turret itself and a foreground abnormal vector group caused by environmental glare interference, and only selecting a geometric center value of the background motion vector group for inter-frame motion compensation to output a stable image sequence.

[0060] Optionally, step 105 can specifically include the following steps: 1051. dividing the set of motion vectors into a high weight group and a low weight group according to the texture weight factor, the high weight group containing motion vectors with a texture weight factor higher than a preset separation threshold, and the low weight group containing the rest of the motion vectors; 1052. calculating a direction consistency density of the motion vectors in the high weight group, classifying motion vectors with the direction consistency density falling into a density peak value region as a background motion vector group, and classifying the low weight group and motion vectors in the high weight group not falling into the density peak value region as a foreground abnormal vector group; 1053. only selecting a geometric center value of all vectors in the background motion vector group as a translation reference to perform reverse displacement compensation on a next frame of image to output a stable image sequence.

[0061] In the above scheme, the separation threshold is a critical weight value for distinguishing the reliability of motion vectors. The direction consistency density is the number of concentrated motion vectors in a specific angle range. The density peak value region is a continuous angle interval with the highest direction consistency density. The geometric center value is the average value of the displacement amounts of all background motion vectors. The reverse displacement compensation is to move the entire image by an amount of displacement opposite to the average displacement.

[0062] In the embodiments of the present application, first, a weight separation threshold is preset through step 1051, and the set of motion vectors is scanned record by record in the format of "ID, horizontal displacement, vertical displacement, weight". When the weight value is greater than or equal to the threshold, the vector is moved into the high weight group, and the rest is moved into the low weight group, generating two independent vector sets. For example, scanning the records "MV102, 3, 3, 1.35; MV205, 1, 0, 0.32"→threshold 1.0→high weight group records MV102, and low weight group records MV205.

[0063] Subsequently, through step 1052, direction binning statistics: 0-360 degrees are divided into 36 direction bins at intervals of 10 degrees, traverse high weight group vectors: calculate the direction angle of each vector, the inverse tangent value of horizontal displacement and vertical displacement; add the vector count to the corresponding direction bin, such as 45.2° into the 40-50° bin; density peak detection: scan the vector quantity of the 36 direction bins, find the highest continuous bin group of the count, for example, 40-50° bin 12, 50-60° bin 11→continuous peak area 40-60°. Group separation: vectors in the peak area of the high weight group→background group; non-peak area vectors of the high weight group+all vectors of the low weight group→foreground group. For example, the high weight group contains 20 vectors, of which 15 fall into the 40-60° peak area→the background group collects 15, and the remaining 5 are combined with the 10 vectors of the low weight group as the foreground group.

[0064] Finally, through step 1053, background group displacement averaging: calculate the average value of the horizontal displacement of all background motion vectors and the average value of the vertical displacement, for example, 10 vector horizontal displacements and 30 pixels→average value 3 pixels. Reverse translation execution: shift the entire next frame image to the left by the average value of pixels and to the top by the average value of pixels. Sequence output: the shifted image is added to the stable video stream, and the original coordinates are superimposed with the displacement compensation value.

[0065] In actual application, in night road monitoring, through step 1051, input the motion vector set: "MV101, 3.1, 3.0, 1.38; MV102, 2.9, 3.2, 1.35; MV201, -1, 4, 0.28; MV301, 0.5, 0.3, 0.41"; set the separation threshold to 1.0, high weight group: MV101 (1.38), MV102 (1.35); low weight group: MV201 (0.28), MV301 (0.41).

[0066] Through step 1052, high weight group vector direction calculation: MV101: displacement (3.1, 3.0)→direction 44.1°; MV102: displacement (2.9, 3.2)→direction 47.8°; direction binning statistics, 10° / bin: 40-50° bin: 2 vectors; other bins: 0; density peak area: 40-50°, the only continuous bin group. Group separation: background group: MV101, MV102, both belong to the 40-50° bin; foreground group: MV201 direction 104°, MV301 direction 31°.

[0067] Through step 1053, the background group displacement averages: horizontal average = (3.1 + 2.9) / 2 = 3.0 pixels; vertical average = (3.0 + 3.2) / 2 = 3.1 pixels; reverse compensation: the next frame image is shifted 3.0 pixels to the left and 3.1 pixels up; output verification: the compensation results in a fluctuation range of the license plate center coordinates of ±0.2 pixels, compared to the original ±3.5 pixels; there is no ghosting around the edges of the license plate characters. The overall solution in step 105, through texture weight layering and directional density peak analysis, accurately separates the actual motion components of the turntable in low-light environments. Image compensation is performed based on a physically consistent motion average, completely eliminating image jitter caused by glare artifacts during nighttime surveillance and outputting a stable, usable video sequence.

[0068] The following is a complete embodiment of steps 101 to 105: During nighttime road monitoring, in step 101, the unmanned intelligent turntable tracks the suspected vehicle at a pitch rate of 30° / second, simultaneously capturing continuous image frames at a frame rate of 50fps, despite the interference from strong headlight glare. The recognition process involves dividing each image frame into 16×16 pixel cells, calculating the maximum and minimum brightness differences within each cell. Cells with a difference value less than 25 are marked as low-texture cells. Adjacent low-texture cells are merged to create a texture-blurred region, which covers the front vehicle's bumper and the road's reflective tape.

[0069] In step 102, the structure variation between adjacent frames is calculated for the blurred bumper texture area: the absolute value of the pixel brightness difference between the current frame and the frame 0.02 seconds prior is accumulated unit by unit, generating a texture complexity distribution map. The high-value area is 1350 near the license plate, and the low-value area is 280 near the direct headlights. A threshold of 1000 is set, and consecutive high-value cells around the license plate are merged to form a "U"-shaped candidate region of 120×40 pixels. This region clearly preserves the indentations of the license plate characters and the texture of the screw holes.

[0070] In step 103, the candidate area is analyzed for five consecutive frames of pixel brightness fluctuations: the license plate character brightness sequence is [50, 120, 55, 125, 60] → the dominant polarization direction is 0°; the filter transmission axis is rotated to 90°, the current angle is 170° → rotated 80° counterclockwise; the image is collected after filtering: the headlight glare brightness is reduced from 200 to 60; the license plate edge grayscale transition is changed from [30, 160] to [65, 68, 71]; the exposure time is compensated to 1 / 67 second, originally 1 / 100 second.

[0071] By step 104, in the license plate candidate area of the polarized light filtering image: tracking the lower right corner point of the license plate character: the current frame (355, 422) matches the previous frame (352, 419) → displacement (3, 3) → vector 45° direction / 4.2 pixel motion vector (355, 422) binds texture value 1350 → weight = 1350 x 0.001 = 1.35 glare point vector (280, 380) binds texture value 320 → weight 0.32 output set: "MV102,3,3,1.35; MV205,1,0,0.32".

[0072] By step 105, weight stratification: threshold 1.0 → high weight group MV102 (1.35), low weight group MV205 (0.32) direction density statistics: high weight group vector direction 45° & 46.8° → 40-50° box density highest background group: MV102 falls into the peak area; geometric center value: horizontal displacement 3 pixels, vertical displacement 3 pixels reverse compensation: next frame left shift 3 pixels / upper shift 3 pixels output result: license plate center coordinate fluctuation reduces from ±3.5 pixels to ±0.2 pixels.

[0073] In the night road monitoring scene of weak light intensity glare, the scheme realizes high-precision image stabilization of the moving target through the threefold synergistic mechanism of texture anti-interference capability quantification screening, dynamic polarization light filtering enhancement of edge features, and texture weight constraint motion vector clustering, accurately separates the real motion of the turntable from the environmental optical artifacts, and realizes high-precision image stabilization of the moving target. The output image sequence completely eliminates the edge tearing and high-frequency jitter caused by headlight glare, and significantly improves the reliability of suspicious vehicle tracking and identification.

[0074] Figure 3 An embodiment of the present application provides a structural schematic diagram of a real-time image stabilization system for intelligent image processing, as shown in Figure 3 The system comprises: An acquisition module 31 is configured to acquire continuous image frames disturbed by environmental strong glare when an unmanned intelligent turntable performs pitching or rotating motion, and identify a texture blur area caused by light reflection in the continuous image frames. A calculation module 32 is configured to calculate a texture complexity distribution graph in combination with a local structure change amount of an adjacent image frame for the texture blur area, and filter a candidate area with a texture feature retention rate higher than a preset standard based on the texture complexity distribution graph. A processing module 33 is configured to apply polarization light filtering processing to the candidate area, generate a polarization light filtering image by adjusting a light filtering angle to match an incident light polarization direction of a current frame, and the polarization light filtering image has the gradient continuity feature of suppressing glare noise and enhancing edge gradient. The extraction module 34 is configured to extract a motion vector field of the candidate region between adjacent frames based on gradient continuity of the polarization filtering image, and perform spatial superposition of the motion vector field and the texture complexity distribution map to form a motion vector set carrying a texture weight factor; The separation module 35 is configured to separate a background motion vector group generated by self motion of the unmanned intelligent turntable and a foreground abnormal vector group caused by environmental glare interference according to the texture weight factor, and only select a geometric center value of the background motion vector group for inter-frame motion compensation to output a stable image sequence.

[0075] Figure 3 The real-time image stabilization system for intelligent image processing can perform Figure 1 The real-time image stabilization method for intelligent image processing of the embodiment has the implementation principle and technical effects which will not be repeated. The specific operation modes of each module and unit of the real-time image stabilization system for intelligent image processing in the above embodiment have been described in detail in the embodiment of the method, and will not be described in detail here.

[0076] In one possible design, Figure 3 The real-time image stabilization system for intelligent image processing of the embodiment can be implemented as a computing device, such as a computer. Figure 4 As shown in the figure, the computing device can include a storage component 41 and a processing component 42. The storage component 41 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 42.

[0077] The processing component 42 is configured to perform the above Figure 1 The real-time image stabilization method for intelligent image processing of the embodiment.

[0078] The processing component 42 can include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component can also be one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic elements for executing the above method.

[0079] The storage component 41 is configured to store various types of data to support the operation of the terminal. The storage component can be implemented by any type of volatile or nonvolatile storage devices, or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic or optical disk.

[0080] Of course, the computing device can also necessarily include other components, such as an input / output interface, a display component, a communication component, etc.

[0081] The input / output interface provides an interface between the processing component and peripheral interface modules, which can be output devices, input devices, etc.

[0082] The communication component is configured to facilitate wired or wireless communication between the computing device and other devices, etc.

[0083] The computing device can be a physical device or an elastic computing host provided by a cloud computing platform, and the computing device can be a cloud server, and the processing component, the storage component, etc. can be a basic server resource rented or purchased from the cloud computing platform.

[0084] The embodiment of the present application also provides a computer storage medium, which stores a computer program, and the computer program can implement the above-mentioned Figure 1 The embodiment shown in the figure is a real-time image stabilization method for intelligent image processing.

[0085] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-mentioned system, device and unit can refer to the corresponding process in the foregoing method embodiment, which will not be repeated here.

[0086] The device embodiment described above is only schematic, wherein the units described as separate components can or can not be physically separated, and the components displayed as units can or can not be physical units, that is, they can be located in one place, or distributed on multiple network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the embodiment scheme. Those skilled in the art can understand and implement it without creative labor.

[0087] Those skilled in the art can clearly understand the implementation of the various embodiments by means of software and the necessary general hardware platform from the above description of the embodiments, and of course, the embodiments can also be implemented by hardware. Based on such understanding, the above technical solutions, essentially or in other words, the part of the prior art that contributes to the technical solutions can be embodied in the form of a software product. The computer software product can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for causing a computer device (which can be a personal computer, a server, etc.) to execute the methods described in the various embodiments or some parts of the methods.

[0088] Finally, it should be noted that: the above examples are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing examples, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing examples, or make equivalent replacement for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A real-time image stabilization method for intelligent image processing, characterized in that, The application relates to a method for generating a stable image sequence by eliminating the interference of strong ambient light on a continuous image frame. When the unmanned intelligent rotary table performs pitching or rotating motion, a continuous image frame interfered by strong ambient light is synchronously acquired, and a texture blurred area caused by light reflection in the continuous image frame is identified; For the texture blurred area, a texture complexity distribution graph is calculated in combination with the local structure change amount of adjacent image frames, and a candidate area with a texture feature retention rate higher than a preset standard is screened based on the texture complexity distribution graph; Polarization filtering processing is applied to the candidate area, a polarization filtering image is generated by adjusting a filtering angle to match the incident light polarization direction of the current frame, and the polarization filtering image has the gradient continuity features of suppressing glare noise and enhancing edge gradient; Based on the gradient continuity features of the polarization filtering image, a motion vector field of the candidate area between adjacent frames is extracted, and the motion vector field is spatially superimposed with the texture complexity distribution graph to form a motion vector set carrying a texture weight factor; According to the texture weight factor, the motion vector set is hierarchically clustered to separate a background motion vector group generated by the motion of the unmanned intelligent rotary table and a foreground abnormal vector group caused by the interference of ambient light, only the geometric center value of the background motion vector group is selected for inter-frame motion compensation, and a stable image sequence is output.

2. The method of claim 1, wherein, Based on the gradient continuity features of the polarization filtering image, a motion vector field of the candidate area between adjacent frames is extracted, and the motion vector field is spatially superimposed with the texture complexity distribution graph to form a motion vector set carrying a texture weight factor, which comprises: In the candidate area of the polarization filtering image, the displacement trajectory of the corresponding edge points between adjacent frames is tracked along the object edge direction, and a motion vector field is formed by the direction and length of the displacement trajectory; The spatial position of each motion vector in the motion vector field is bound with the value of the texture complexity distribution graph at the same spatial position; According to the bound value, each motion vector is given a texture weight factor, and the texture weight factor is in a proportional relationship with the value of the texture complexity distribution graph, thereby forming a motion vector set carrying a texture weight factor.

3. The method of claim 1, wherein, The polarization filtering processing applied to the candidate area generates a polarization filtering image by adjusting a filtering angle to match the incident light polarization direction of the current frame, and the polarization filtering image has the gradient continuity features of suppressing glare noise and enhancing edge gradient, which comprises: According to the light reflection characteristics of the candidate area, the dominant polarization direction of the incident light in the candidate area is analyzed, and the dominant polarization direction is determined by the periodic change mode of the pixel brightness in the continuous frames; The light transmission axis angle of the physical filter is dynamically rotated to form an orthogonal relationship between the light transmission axis angle and the dominant polarization direction; According to the adjusted light transmission axis angle, the light rays of the candidate area passing through the physical filter are collected to generate a polarization filtering image, and the brightness amplitude of the glare noise in the polarization filtering image is reduced and the light-dark transition continuity of the object edge is improved.

4. The method according to claim 1, wherein The motion vector set is hierarchically clustered according to the texture weight factor, background motion vector groups generated by self-motion of the unmanned intelligent turntable and foreground abnormal vector groups caused by environmental glare interference are separated, only the geometric center value of the background motion vector groups is selected for inter-frame motion compensation, and a stable image sequence is output, including: The motion vector set is divided into a high weight group and a low weight group according to the texture weight factor, the high weight group contains motion vectors with a texture weight factor higher than a preset separation threshold, and the low weight group contains the remaining motion vectors; The direction consistency density of the motion vectors in the high weight group is calculated, the motion vectors falling into the density peak value region are classified as background motion vector groups, and the low weight group and the motion vectors not falling into the density peak value region in the high weight group are classified as foreground abnormal vector groups; Only the geometric center value of all vectors in the background motion vector groups is selected as the translation reference, reverse displacement compensation is performed on the next frame of image, and a stable image sequence is output.

5. The method of claim 1, wherein, For the texture blur area, a texture complexity distribution map is calculated in combination with the local structure change amount of the adjacent image frames, and a candidate area with a texture feature retention rate higher than a preset standard in the texture blur area is selected based on the texture complexity distribution map, including: For each image unit in the texture blur area, the structure change amount of the corresponding image unit between the current frame and the adjacent frame is calculated, and the structure change amount is represented by the cumulative value of the difference amplitude of the pixel values in the unit; The structure change amount is mapped to each image unit position of the texture blur area to generate a texture complexity distribution map, and the value of the texture complexity distribution map reflects the ability strength of the image unit to resist glare interference; In the texture complexity distribution map, image units with a value higher than a preset threshold are marked, and continuously distributed marked units in the image units are combined into a candidate area, and the candidate area covers a continuous spatial range with a texture feature retention rate higher than a preset standard.

6. The method of claim 3, wherein, The light transmission axis angle of the dynamic rotating physical filter is adjusted to form an orthogonal relationship between the light transmission axis angle and the dominant polarization direction, including: The angle value of the dominant polarization direction is measured, and the orthogonal angle gap between the angle value and the light transmission axis angle is calculated; The motor driving pulse number is generated according to the orthogonal angle gap, and the rotating motor is driven to rotate the physical filter by the gap compensation angle; During rotation, when the feedback value of the light transmission axis angle sensor reaches the orthogonal angle of the angle value, the rotating motor is stopped and the light transmission axis is locked.

7. The method of claim 2, wherein, According to the bound values, a texture weight factor is assigned to each motion vector, the texture weight factor is in a positive proportional relationship with the value of the texture complexity distribution map, and a motion vector set carrying the texture weight factor is formed, including: The bound values and the texture complexity distribution map values corresponding to the motion vector spatial positions are taken as the texture credibility values of the current motion vectors; The texture credibility value is input into a linear converter to generate a texture weight factor through a fixed-multiplying amplification operation, and a value of the texture weight factor is equal to the texture credibility value multiplied by a preset proportion coefficient; The texture weight factor is written into an attribute field of a corresponding motion vector to construct a motion vector set carrying the texture weight factor.

8. A real-time image stabilization system for intelligent image processing, characterized in that, The method comprises the following steps: The acquisition module is configured to acquire, synchronously, continuous image frames disturbed by environmental strong glare light when an unmanned intelligent turntable performs pitching or rotating motion, and identify a texture blur area caused by light reflection in the continuous image frames; The calculation module is configured to calculate a texture complexity distribution graph in combination with a local structure change amount of adjacent image frames for the texture blur area, and filter a candidate area with a texture feature retention rate higher than a preset standard based on the texture complexity distribution graph; The processing module is configured to apply a polarized light filtering process to the candidate area, generate a polarized light filtering image by adjusting a filtering angle to match an incident light polarization direction of a current frame, and the polarized light filtering image has both glare noise suppression and gradient continuity features of enhanced edge gradient; The extraction module is configured to extract a motion vector field of the candidate area between adjacent frames based on the gradient continuity features of the polarized light filtering image, and perform spatial superposition of the motion vector field and the texture complexity distribution graph to form a motion vector set carrying a texture weight factor; The separation module is configured to perform hierarchical clustering on the motion vector set according to the texture weight factor, separate a background motion vector group generated by the motion of the unmanned intelligent turntable itself and a foreground abnormal vector group caused by environmental glare light disturbance, select only a geometric center value of the background motion vector group for inter-frame motion compensation, and output a stable image sequence.

9. A computing device, comprising: The method comprises a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement the real-time image stabilization method for intelligent image processing according to any one of claims 1 to 7.

10. A computer storage medium, characterized in that, The computer program is stored in a computer and is executed by the computer to implement the real-time image stabilization method for intelligent image processing according to any one of claims 1 to 7.

Images