Impact deformation feature extraction method based on high-speed camera shooting and image processing

Abstract

The invention provides an impact deformation feature extraction method based on high-speed camera shooting and image processing, and the method comprises the steps: collecting the transient deformation data of the edge package of a remote controller housing under impact through high-speed camera shooting, recording the change speed of a strain rate, obtaining the material characteristics, the size of a package structure, and the geometric discontinuous features of an opening and closing part, generating continuous frame sequence images; performing strain field data quantification processing on the third visualization result, extracting strain field characteristic data, and judging influence weights of stress wave reflection on strain fields of different functional areas of the shell in combination with a falling height and an impact angle of the remote controller; and improving edge packaging design parameters according to the deformation buffer effect and discontinuous quantitative indexes at the opening and closing positions, verifying the improvement amplitude of the deformation buffer effect, and obtaining strain field distribution data of the optimized shell.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of information technology, and in particular to an impact deformation feature extraction method based on high-speed photography and image processing. BACKGROUND

[0002] In the field of modern industrial design and manufacturing, the study of the impact resistance of product casings is of great significance, especially in the protection of electronic devices, which directly affects the service life of the product and the user experience. How to evaluate and optimize the design of the casing through scientific methods to cope with various accidental impacts has become a key issue that cannot be ignored in the industry. The research in this field not only affects product reliability, but also plays an important role in improving market competitiveness. However, the current analysis methods for the impact resistance of casings still have obvious shortcomings. Many traditional methods often rely on static testing or simple simulation, which is difficult to capture the dynamic changes in the impact moment and cannot accurately reveal the subtle deformation characteristics of the casing under complex stress. This limitation leads to a lack of sufficient data support for design optimization, affecting the improvement of the protection capability of the product. In-depth, the core challenge in this field is mainly concentrated on the accurate extraction and analysis of dynamic deformation characteristics during the impact process. First of all, due to the complex and variable deformation buffer effect of the casing edge during the impact, existing technologies are difficult to capture its subtle changes in real time, leading to an insufficient understanding of the protection mechanism. This deficiency further leads to difficulties in analyzing stress wave propagation and interference patterns, as the reflection and superposition of stress waves in the edge area significantly affect the overall structure response, but there is a lack of effective observation means. More importantly, the deformation discontinuity problem in the edge design of the opening and closing directly affects the stability of the overall casing, which becomes a technical difficulty that needs to be overcome. These challenges are layered and together constitute an obstacle to the comprehensive evaluation of the impact resistance of the casing. Therefore, how to dynamically extract impact deformation characteristics through innovative methods and accurately analyze the adjustment effect of the edge packaging design on the impact response has become a key problem in improving the evaluation of the impact resistance of the casing. SUMMARY

[0003] The present application provides an impact deformation feature extraction method based on high-speed photography and image processing, mainly including:

[0004] Collecting transient deformation data of the edge packaging of the remote controller casing under impact through high-speed photography, recording the strain rate change speed, obtaining material properties, packaging structure size and opening and closing geometric discontinuity characteristics, and generating a continuous frame sequence image;

[0005] Denoising and edge enhancement are performed on the continuous frame sequence image to extract stress wave reflection and deformation buffer effect characteristics, identify the unique thin-walled structure vibration mode of the remote controller casing and the stress concentration area of the battery compartment edge, and obtain a first set of processed images;

[0006] Based on the first set of processed images, a strain field visualization model is constructed. The deformation buffer effect characteristics, stress wave reflection, shell thin wall thickness distribution and stiffness non-uniformity caused by button hole positions are input into the strain field visualization model. The actual geometric features of the remote control shell are mapped through finite element mesh, and the strain gradient distribution considering the shell structural characteristics is output.

[0007] The initial parameters of the color mapping range are determined based on the material properties. The strain gradient distribution is judged to conform to the preset distribution range by combining the gradient display accuracy threshold. The strain field visualization model is iteratively optimized based on the geometric discontinuity characteristics of the opening and closing points in real time, and a second visualization result reflecting the influence of discontinuity at the opening and closing points is generated.

[0008] Frequency domain analysis is performed on the second visualization result to obtain the interference mode resolution of the stress wave reflection in the thin-walled shell structure. If it is lower than the preset threshold, the local contrast of the stress concentration area and the edge discontinuity position is enhanced, and the gradient display accuracy is adjusted according to the strain rate change to obtain the third visualization result.

[0009] The strain field data of the third visualization result is quantified, the strain field feature data is extracted, and the influence weight of stress wave reflection on the strain field of different functional areas of the shell is determined by combining the drop height and impact angle of the remote control.

[0010] The edge packaging design parameters were improved based on the quantitative indicators of deformation buffer effect and discontinuity at the opening and closing points. The improvement of deformation buffer effect was verified, and the strain field distribution data of the optimized shell were obtained.

[0011] If the strain gradient at the opening and closing points in the optimized strain field distribution data exceeds the threshold, the impact data of the improved scheme is re-acquired, the color mapping parameters are adjusted, and a fourth visualization result is obtained. This result is then compared with the initial model to obtain a design improvement effect evaluation report.

[0012] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects:

[0013] This invention discloses a method for extracting impact deformation features based on high-speed imaging and image processing. It acquires transient deformation data of a housing under impact using a high-speed camera, and constructs a strain field visualization model by combining material properties and structural parameters. The method processes the acquired images to extract stress wave reflection and deformation buffering effect features, identifying vibration modes and stress concentration regions of thin-walled structures. It maps actual geometric features through a finite element mesh, outputting a strain gradient distribution considering structural characteristics. Based on the discontinuities at openings and closing points, iteratively optimizes the model, generating visualization results reflecting their impact. Frequency domain analysis and data quantization are performed on the results to extract feature data and determine the influence of stress wave reflection. Finally, the packaging design is improved based on quantified indicators, verifying the optimization effect and achieving accurate evaluation and effective improvement of the impact resistance performance of the remote control housing. Attached Figure Description

[0014] Fig. 1 This is a flowchart of an impact deformation feature extraction method based on high-speed photography and image processing according to the present invention.

[0015] Fig. 2 This is a schematic diagram of an impact deformation feature extraction method based on high-speed photography and image processing according to the present invention. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0017] like Figs. 1-2 This embodiment of an impact deformation feature extraction method based on high-speed photography and image processing may specifically include:

[0018] S101. Acquire transient deformation data of the remote control housing edge encapsulated under impact using a high-speed camera, record the strain rate change rate, obtain material properties, encapsulation structure dimensions and geometric discontinuities at the opening and closing points, and generate a continuous frame sequence of images.

[0019] A high-speed camera captures the dynamic response of the remote control housing edge under impact load at a preset acquisition frequency, obtaining displacement time-history curves of each monitoring point on the housing edge. The instantaneous deformation velocity of each point is calculated by dividing the displacement difference between adjacent frames by the time interval. Strain rate distribution data is obtained based on the ratio of deformation velocity to the original edge size, forming a continuous frame sequence image containing timestamps and spatial coordinates. Peak regions of strain rate distribution data are extracted from the continuous frame sequence image. Gray-scale gradient analysis is performed on the image at the corresponding time of the peak region. The stress distribution of the edge structure is calculated by tracking the displacement of pixels. Combined with the stress and strain curve of ABS material, when the stress value reaches a preset proportional threshold of the material's yield strength, the region is determined to have entered the plastic deformation stage, and the critical strain rate value and corresponding spatial coordinates at this time are recorded. Based on the critical strain rate value and its corresponding spatial coordinates, combined with the snap-fit ​​clearance parameters, the stress amplification factor at each snap-fit ​​location is calculated using the stress concentration factor K = σmax / σ0, where σmax is the local maximum stress and σ0 is the nominal stress. If the instantaneous strain rate at a certain location exceeds the critical value and the duration is greater than a preset time threshold, it is determined that the snap-fit ​​at that location is at risk of detachment, and the location with the largest stress concentration factor K value is identified as the weak point. For the stress concentration factor K value at the weak point location, combined with the geometric parameters of the battery compartment edge chamfer radius R and sealing groove depth H, the crack propagation trend is calculated using the stress intensity factor KI = Yσ√πa, where Y is the shape factor, σ is the applied stress, and a is the crack length. Based on the changes in KI value under different impact angles, a dataset of failure critical conditions for the shell edge encapsulation structure is obtained, generating a continuous frame sequence of images labeled with material property parameters, structural dimension parameters, and geometric discontinuities.

[0020] Specifically, the application of high-speed camera technology in the impact test of remote control housing involves the coordinated efforts of several key aspects.

[0021] In one possible implementation, the choice of acquisition frequency for the high-speed camera directly affects the accuracy of strain rate calculation. When an impact occurs, the deformation process at the shell edge is extremely brief, typically completed within milliseconds, thus requiring extremely high temporal resolution to capture this transient process. By arranging multiple monitoring points at the shell edge to form a dense measurement grid, the displacement changes at each monitoring point are precisely recorded, and this data forms the basis for subsequent analysis.

[0022] Specifically, the acquisition of strain rate distribution data involves a gradual transformation from displacement to velocity and then to strain rate. The time interval between two adjacent frames is extremely short; by comparing the positional changes of the same monitoring point in adjacent frames, the instantaneous velocity at that point can be calculated. Dividing this velocity value by the original size of that location yields the local strain rate. The advantage of this calculation method is that it can reflect the true response characteristics of the material under high-speed loading conditions, which is crucial information that traditional static testing methods cannot obtain.

[0023] It's important to note that grayscale gradient analysis plays a crucial role in processing continuous frame sequences of images. When a region in an image deforms, its grayscale value changes accordingly. By calculating the grayscale gradient, the boundaries and extent of the deformation can be identified. This method is particularly suitable for identifying regions of concentrated strain, as these areas often exhibit abrupt changes in grayscale gradient. Combined with the material's stress-strain curves, when the stress in a region is detected to be approaching its yield strength, it indicates that the region is about to enter an irreversible plastic deformation stage, which is critical for assessing the safety of the shell.

[0024] In one embodiment, the calculation of the stress concentration factor reveals the distribution pattern of weak points in the structure. Due to abrupt changes in geometry, snap-fit ​​locations often become high-incidence areas of stress concentration. By calculating the ratio of the local maximum stress to the nominal stress, the degree of danger at each location can be quantitatively assessed. The technical advantage of this assessment method is that it can predict potential failure locations before failure occurs, providing clear directions for structural optimization.

[0025] Preferably, the introduction of the stress intensity factor makes it possible to predict the crack propagation trend. The value of the shape factor Y depends on the crack geometry and loading method. For thin-walled structures such as remote control housings, edge cracks and surface cracks are the most common failure modes.

[0026] S102. Denoise and enhance the edges of the continuous frame sequence images, extract the features of stress wave reflection and deformation buffer effect, identify the vibration mode of the thin-walled structure unique to the remote control shell and the stress concentration area at the edge of the battery compartment, and obtain the first processed image set.

[0027] Median filtering is applied to remove salt-and-pepper noise from a series of consecutive frame images. The median value of the center pixel is replaced by the median value of the surrounding preset neighborhood of each pixel to obtain a denoised image. Then, the second derivative of each pixel in the denoised image is calculated using the Laplacian operator. This derivative value is multiplied by an enhancement coefficient and then superimposed onto the corresponding pixel value in the denoised image to obtain an edge-enhanced image. Regions with abrupt grayscale changes are extracted from the edge-enhanced image. The propagation path of stress waves is identified by calculating the grayscale difference between adjacent pixels. When the grayscale difference exceeds a preset threshold, the coordinates and timestamp of that location are recorded. The wave velocity is obtained by dividing the distance between adjacent frames from the peak stress wave location by the time interval. The wave velocity attenuation rate is calculated by comparing the initial wave velocity with the current wave velocity. If the wave velocity attenuation rate is lower than a preset value, a deformation buffering effect is determined to exist in that region, thus obtaining stress wave reflection characteristic data and a buffering effect distribution map. For the time-domain signal in the stress wave reflection characteristic data, it is converted into a frequency-domain signal through discrete Fourier transform. Each frequency component and its corresponding amplitude value are extracted, and the arithmetic mean of the amplitudes of all frequency components is calculated as the average amplitude. When the amplitude of a certain frequency component exceeds a preset multiple of the average amplitude, the vibration mode corresponding to that frequency is determined as the dominant mode of the thin-walled structure, and the frequency value and spatial mode shape distribution of the dominant mode are recorded. Based on the region with the largest amplitude in the spatial mode shape distribution of the dominant mode, combined with the weak points in the buffer effect distribution map, the stress gradient value is obtained by calculating the difference in stress values ​​between adjacent locations and dividing by the distance between locations. When the stress gradient value of a certain region is greater than a preset multiple of the average gradient value of the surrounding region, this region is marked as a stress concentration area at the edge of the battery compartment. The edge enhancement image, stress wave characteristics, dominant mode distribution, and stress concentration area annotation information are integrated to generate the first processed image set.

[0028] For example, median filtering is a crucial step in analyzing consecutive frame sequences of images. Its principle is to replace the center pixel's value with the median value of its surrounding neighborhood, effectively removing salt-and-pepper noise. Suppose that in a set of images, a pixel has a grayscale value of 255, while the grayscale values ​​of its eight surrounding pixels are mostly around 120. Median filtering will replace this pixel's value with a median value close to 120, smoothing out noise interference. This method is particularly suitable for images captured by high-speed cameras, as these images often introduce noise due to equipment shake or changes in lighting.

[0029] For example, in edge enhancement processing, the Laplacian operator is used to highlight areas in an image where grayscale values ​​change drastically, i.e., potential stress concentration points. Its principle is to identify edges by calculating the second derivative of each pixel; the larger the derivative value, the more pronounced the grayscale change. Suppose that in a frame of an image, the derivative value of a certain region reaches 5, while the surrounding region is only around 1. After multiplying by an enhancement factor of 2, the pixel value of that region will be significantly improved, forming clearer edge features, facilitating subsequent analysis.

[0030] For example, grayscale difference calculation is a core method in extracting stress wave propagation paths. Assuming adjacent pixel grayscale values ​​are 100 and 150 respectively, with a difference of 50, and a preset threshold of 30, this location is recorded as a stress wave propagation path point, and the wave velocity is calculated using a timestamp. If the initial wave velocity is 500 m / s and the current wave velocity is 400 m / s, the ratio is 0.8. If the preset attenuation rate is 0.85, then a deformation buffering effect is determined to exist in this region. This analysis helps identify the energy absorption characteristics of the shell material.

[0031] For example, in frequency domain signal analysis, the Discrete Fourier Transform converts a time-domain signal into a frequency-domain signal, extracting the vibration frequency and amplitude. Assuming an average amplitude of 10, and an amplitude of 25 at a certain frequency, exceeding a preset multiple of 2, this frequency is identified as the dominant mode, and its frequency value is recorded, such as 50 Hz. This method can reveal the main vibration modes of thin-walled structures under impact, providing a reference for shell design.

[0032] For example, in stress gradient calculation, assuming a stress difference of 20 Pa in a certain area, a location spacing of 0.002 meters, and a stress gradient of 10,000 Pa / m, if the surrounding average gradient is 5,000 Pa / m, exceeding a preset multiple of 1.5 times, it is marked as a stress concentration area at the edge of the battery compartment. This marking, combined with edge enhancement images and stress wave characteristics, can more comprehensively present the distribution of weak points in the shell.

[0033] For example, when integrating the first set of processed images, edge-enhanced images, stress wave features, and stress concentration area annotation information are overlaid to form an intuitive visual presentation. If, in the image set, a certain buckle location simultaneously displays a high grayscale difference and a stress gradient peak, then that area is highlighted as a potential failure point.

[0034] S103. Construct a strain field visualization model based on the first processed image set. Input the deformation buffer effect characteristics, stress wave reflection, shell thin wall thickness distribution and stiffness non-uniformity caused by button hole position into the strain field visualization model. Map the actual geometric features of the remote control shell through finite element mesh and output the strain gradient distribution considering the shell structural characteristics.

[0035] Pixel coordinates and grayscale values ​​of stress concentration regions are extracted from the first processed image set. A strain field data matrix is ​​constructed based on the ratio of strain to image grayscale in mechanics of materials, ε = k × G, where ε is the strain value, k is the calibration coefficient, and G is the normalized grayscale value. Strain values ​​between adjacent pixels are calculated using bilinear interpolation. The wave velocity attenuation rate in the deformation buffer effect feature is then converted into a buffer coefficient and superimposed onto the corresponding position to obtain the initial strain field of the fused buffer characteristic. For the initial strain field of the fused buffer characteristic, boundary condition parameters are set based on the spatial location of the stress wave reflection points. The thin-wall thickness distribution of the shell is calculated by comparing the difference between the standard wall thickness and the measured pixel grayscale. After identifying the keyhole contour, the stiffness reduction coefficient Kr = Ir / I0 is calculated using the reduction ratio of the cross-sectional moment of inertia, where Ir is the actual cross-sectional moment of inertia and I0 is the cross-sectional moment of inertia without holes. If Kr is lower than a preset threshold, it is marked as a region with non-uniform stiffness. A structural feature dataset containing boundary conditions, material thickness, and stiffness distribution is generated. Based on the structural feature dataset, a strain field visualization model based on the finite element principle was established. Four-node quadrilateral elements were used to mesh the shell surface. The mesh density was increased to a preset multiple of the standard density at the rounded corners. The mesh arrangement was adjusted along the stiffener direction, and fixed constraints were set at the screw pillar locations. The initial strain field data was mapped to the mesh nodes using a first-order Lagrangian function N. The strain value εi = ΣNjεj at node j was calculated, and the strain gradient was obtained by dividing the strain difference between adjacent nodes by the spacing. The output shows the strain gradient distribution considering the structural characteristics of the shell.

[0036] In one possible implementation, establishing the proportional relationship between strain and image grayscale involves the practical application of materials mechanics principles. When the remote control housing is subjected to an impact load, the material deforms, leading to a change in the intensity of reflected light, which is then reflected as a change in grayscale value in the image. The calibration coefficient k is determined through a pre-conducting tensile test. Under known stress conditions, the strain value of the material is measured, and the corresponding image grayscale change is recorded. A linear relationship coefficient is obtained by fitting the data using the least squares method. Normalization processing makes the measurement results comparable under different lighting conditions, improving the accuracy of the strain field construction.

[0037] Specifically, bilinear interpolation plays a crucial role in strain field data processing. Due to the discrete nature of image acquisition, data gaps exist between adjacent pixels, leading to a discontinuous, stepped distribution in the strain field if used directly. Bilinear interpolation, by considering the contributions of four known data points surrounding the target point and performing a weighted average based on distance, ensures a smooth transition in the strain field. This processing method is particularly suitable for curved structures such as remote control housings, accurately reflecting the continuous distribution characteristics of strain.

[0038] It should be noted that the conversion of wave velocity attenuation rate to buffer coefficient is based on the principle of energy dissipation. When a stress wave propagates through a material, some energy is absorbed by the damping effect within the material, resulting in a gradual decrease in wave velocity. The greater the attenuation rate, the stronger the buffering capacity of that region. By establishing the correspondence between attenuation rate and buffer coefficient, the absorption capacity of different regions for impact energy can be quantitatively assessed, which is of great significance for identifying weak points in the shell.

[0039] In one embodiment, the calculation of the moment of inertia of the cross section directly reflects the change in structural stiffness. The presence of the keyhole creates a gap in the originally continuous shell material, resulting in a decrease in the area's resistance to bending deformation. By calculating the ratio of the moment of inertia of the perforated section to the non-perforated section, the degree of stiffness loss can be quantitatively assessed. When this ratio is below 0.7, it indicates a significant reduction in the load-bearing capacity of the area, requiring special attention in the design.

[0040] Preferably, the four-node quadrilateral element is chosen based on its advantages in thin-shell structure analysis. This element type can adapt well to curved surface geometries while maintaining high computational efficiency. The mesh density is increased in the rounded corner transition regions because these locations exhibit geometric abrupt changes and significant stress concentration, requiring a finer mesh to accurately capture the abrupt changes in strain gradient. The first-order Lagrangian function provides interpolation rules from element nodes to arbitrary points within the element, enabling the transformation of discrete nodal data into a continuous strain field distribution, thus achieving accurate calculation of the strain gradient.

[0041] S104. Determine the initial parameters of the color mapping range based on the material properties, and determine whether the strain gradient distribution conforms to the preset distribution range by combining the gradient display accuracy threshold. Iteratively optimize the strain field visualization model based on the geometric discontinuity characteristics of the opening and closing points in real time, and generate a second visualization result that reflects the influence of discontinuity at the opening and closing points.

[0042] Based on the yield strain and fracture strain values ​​of ABS material, minimum and maximum values ​​of color mapping are set as initial parameters. The strain gradient distribution data is mapped to the RGB color space according to a linear proportional relationship, where blue corresponds to the low strain region and red corresponds to the high strain region. By calculating the Euclidean distance between the RGB values ​​of adjacent mesh elements, if the distance is less than the preset gradient display accuracy threshold, the color mapping range is expanded to enhance contrast, resulting in a preliminary visualization image of the strain field. The geometric boundary positions of the opening and closing points are identified from the preliminary visualization image of the strain field. The width of the buckle fit gap and the misalignment distance of the battery compartment cover edge are measured. The geometric discontinuity level is determined based on the ratio of the misalignment distance to the shell thickness. When the discontinuity level exceeds the preset level, displacement constraint boundary conditions are added at that position. Local strain field correction data considering the influence of the opening and closing points are obtained through finite element re-analysis. For the region of difference between the corrected strain field data and the original strain field data, the grid node density and shape function order in this region are adjusted to enable the corrected data to be accurately mapped into the visualization image. The nonlinear parameters of the color mapping function are iteratively adjusted. The iteration stops when the pixel difference rate of the images generated by two adjacent iterations is less than the preset convergence threshold. The optimized color mapping is applied to the entire strain field to generate a second visualization result that reflects the discontinuity at the opening and closing points.

[0043] In one possible implementation, the color mapping range is determined based on a combination of material mechanical properties and visual perception. The yield strain of ABS material is typically between 0.02 and 0.04, while the fracture strain can reach over 0.1. This range determines the boundaries of the color mapping. When linearly mapping strain values ​​to the RGB color space, using a contrast between warm and cool tones can intuitively reflect the degree of danger of the stress state. Blue represents the safe zone, gradually transitioning to green and yellow as strain increases, eventually reaching the high-risk red zone. This color gradation aligns with the human eye's intuitive perception of danger levels.

[0044] Specifically, Euclidean distance calculation plays a crucial role in assessing color differences. The RGB values ​​of adjacent grid cells form two three-dimensional vectors. By calculating the distance between these two vectors in the color space, the smoothness of color transitions can be quantitatively evaluated. When the distance value is too small, it indicates insufficient color differentiation between adjacent areas, making it difficult for the observer to discern subtle changes in the strain gradient. By expanding the mapping range, the same strain difference will correspond to a larger color difference, thereby enhancing the recognizability of the visualization.

[0045] It should be noted that the classification of geometric discontinuity levels reflects the degree of damage to structural integrity. Misalignment at the opening and closing points of the remote control housing not only affects the appearance but, more importantly, alters the stress transmission path. The ratio of misalignment distance to housing thickness provides a dimensionless parameter, allowing for a unified evaluation standard for remote control products of different sizes. When this ratio exceeds 0.1, it indicates that the deformation at the opening and closing point has significantly affected the stress distribution in the surrounding area, requiring special treatment.

[0046] In one embodiment, the application of displacement constraint boundary conditions simulates the constraint effect in actual use. The battery compartment cover is connected to the main housing by a snap-fit ​​mechanism, and upon impact, a local constraint force is generated at the snap-fit ​​location. This constraint force limits the relative displacement at the opening and closing points, leading to stress redistribution. Through finite element reanalysis, the true strain field considering this constraint effect can be calculated, and this approach is closer to reality than the simple continuum assumption.

[0047] Preferably, adjusting the grid node density and shape function order achieves a balance between computational accuracy and efficiency. In regions with drastic strain gradient changes, using a denser grid and higher-order shape functions can accurately capture local features. Conversely, in regions with gentle strain distribution, maintaining a lower grid density reduces computational cost. This adaptive grid strategy makes the entire visualization process both accurate and efficient. The convergence criterion during the iterative optimization process ensures the stability of the visualization results. By comparing the pixel difference rate of images generated in two adjacent iterations, it can be determined whether the optimization has reached a stable state.

[0048] S105. Perform frequency domain analysis on the second visualization result to obtain the interference mode resolution of the stress wave reflection in the thin-walled shell structure. If it is lower than the preset threshold, then enhance the local contrast of the stress concentration area and the edge discontinuity position, and adjust the gradient display accuracy according to the strain rate change to obtain the third visualization result.

[0049] A two-dimensional fast Fourier transform is performed on the second visualization result to convert the strain distribution image in the spatial domain into a frequency domain representation. By calculating the energy ratio of high-frequency components to low-frequency components in the spectrum, interference fringes formed by multiple reflections of stress waves in the thin-walled structure are identified. The frequency domain distance between adjacent fringe peaks is measured, and its reciprocal is taken as the interference mode resolution value. If this value is lower than a preset threshold, the corresponding spatial domain location is recorded as a region with insufficient resolution. For the corresponding location of the region with insufficient resolution in the spatial domain, the coordinates of stress concentration points and edge discontinuities are identified. Histogram stretching is performed on these local regions to expand the original grayscale range to a wider interval. The increase factor of the grayscale difference between adjacent pixels before and after processing is calculated as the contrast enhancement coefficient. This coefficient is multiplied by the original color value and remapped to obtain a locally contrast-enhanced image. The strain rate numerical sequence of each grid node is extracted from the local contrast-enhanced image. The strain rate change rate is calculated by dividing the difference in strain rate between adjacent time points by the time interval. When the change rate of a node exceeds a preset threshold, the color level of the node and its neighborhood is increased from the standard level to a preset multiple, making the color transition of the strain gradient more delicate. All adjusted display parameters are integrated to generate a third visualization result.

[0050] In one possible implementation, the two-dimensional Fast Fourier Transform (FFT) plays a unique role in strain field visualization analysis. When stress waves propagate and reflect within the thin-walled structure of a remote control, they form a fringe pattern similar to optical interference. These fringes exhibit a periodic distribution of alternating bright and dark areas in the spatial domain. After being transformed to the frequency domain via Fourier transform, the periodic characteristics of the fringes form distinct peaks in the spectrum. High-frequency components represent rapidly changing details, while low-frequency components correspond to slowly changing overall trends. The energy ratio between the two reflects the richness of detail in the image, which directly relates to the clarity of the interference pattern.

[0051] Specifically, measuring the spacing between interference fringes involves a frequency domain peak identification process. In a spectrum diagram, each peak corresponds to a specific spatial frequency, and the frequency domain distance between adjacent peaks represents the repetition period of the fringes in the spatial domain. Taking the reciprocal converts the frequency to a spatial scale, and the resulting value represents the number of resolvable fringes per unit length, i.e., the resolution. When the remote control housing material is thin or there is severe stress concentration, the interference fringes become denser. If this exceeds the display capability of the visualization system, insufficient resolution will occur.

[0052] It's worth noting that histogram stretching is remarkably effective at enhancing local contrast. In the original image, grayscale values ​​in stress concentration areas and discontinuous edges may be concentrated within a narrow range, making it visually difficult to distinguish different stress levels. By statistically analyzing the grayscale distribution of these local areas, identifying the actual minimum and maximum values, and then linearly mapping this range to a wider display area, the previously similar grayscale values ​​appear more distinct. This processing is particularly suitable for locations with abrupt geometric changes, such as the opening and closing points of remote controls, because these areas experience dramatic stress gradient changes and require higher display contrast for accurate representation.

[0053] In one embodiment, the calculation of the strain rate change rate reveals the dynamic characteristics of the material response. When a remote control is subjected to impact, the strain rate at different locations is not constant but changes rapidly over time. By recording the strain rate values ​​at each moment, calculating the difference between adjacent moments and dividing by the time interval, the time derivative of the strain rate is obtained. The larger this derivative value, the more drastic the change in stress state at that location, and the higher the corresponding risk of material damage.

[0054] Preferably, the dynamic adjustment of the color levels enables adaptive optimization of visualization accuracy. A standard color map may use 256 gray levels, but in regions where strain rate changes rapidly, this accuracy may be insufficient to represent subtle gradient changes. By increasing the color levels to higher levels, such as 1024 or more, richer strain information can be presented within a limited display space.

[0055] S106. Perform strain field data quantification on the third visualization result, extract strain field feature data, and determine the influence weight of stress wave reflection on the strain field of different functional areas of the shell by combining the drop height and impact angle of the remote control.

[0056] The pixel grayscale values ​​in the third visualization result are reverse-mapped. Based on the preset grayscale-strain conversion relationship, the strain values ​​of each grid node are obtained. The coordinates of the latch position are identified, and the maximum strain value within that area is extracted. The strain gradient steepness is obtained by calculating the ratio of the strain difference between adjacent nodes at the battery compartment edge to the distance. The ratio of the maximum stress around the button hole to the far-field stress is measured as the stress concentration factor. The overall deformation of the shell is calculated and accumulated through the integral of the strain-displacement relationship, forming a feature data set containing four quantitative indicators. Based on the maximum strain value of the latch and the gradient steepness of the battery compartment edge in the feature data set, combined with the drop height recorded by the high-speed camera, the impact energy is calculated by multiplying the mass and height. The impact angle is determined by the angle between the shell surface normal and the impact direction at the moment of impact. If the angle is less than a preset threshold, it is determined that a frontal impact dominates the strain in the latch area; otherwise, it is determined that a side impact dominates the strain at the battery compartment edge, thus obtaining the strain distribution characteristics under different impact modes. Based on the strain distribution characteristics under different impact modes, the transmittance and reflectance of stress waves at each interface are calculated according to the material impedance difference. The incident strain value is multiplied by the reflectance to obtain the reflected strain component. The incident and reflected strain components of each functional area are superimposed to obtain the total strain level. The product of the total strain level and the corresponding index in the characteristic data group of the area is normalized, and the influence weight values ​​of stress wave reflection on the strain field of four functional areas—the snap-fit ​​area, the edge of the battery compartment, the area around the button hole, and the overall shell—are output.

[0057] In one possible implementation, the inverse mapping process of grayscale to strain conversion is based on fundamental principles of materials mechanics. Each pixel in the third visualization carries specific color information, which is derived from the strain value. The inverse mapping reverses this process, restoring the corresponding strain value from the RGB values. The accuracy of this conversion directly affects the reliability of subsequent feature extraction, especially when identifying the strain state of key structural components.

[0058] Specifically, the calculation of the strain gradient steepness reflects the spatial rate of change of the strain field. At locations with abrupt geometric changes, such as the edge of the battery compartment, discontinuities in material thickness can lead to drastic changes in strain distribution. By measuring the strain difference between adjacent nodes and dividing it by the node spacing, the resulting gradient value can quantitatively describe the severity of this change. A larger gradient indicates more severe stress concentration at that location, and a higher risk of material failure. This parameter is of great significance for assessing structural weaknesses in the remote control housing.

[0059] It should be noted that the application of the strain-displacement integral in calculating the overall deformation involves the principles of continuum mechanics. Strain represents the deformation per unit length, and by performing path integration on the strain field, the displacement value at each point can be obtained. By vector superimposing the displacements of all nodes and considering directionality, the total deformation reflects the overall response of the shell under impact load. This overall index, combined with local strain parameters, can comprehensively assess the safety of the structure.

[0060] In one embodiment, the calculation of impact energy reveals the intrinsic link between drop height and structural damage. When a remote control is dropped from a specific height, its potential energy is converted into kinetic energy, which is ultimately transformed into impact energy that deforms the casing at the moment of impact. The product of mass and height provides a measure of potential energy, and this energy value determines the magnitude of the impact load. The determination of the impact angle further refines the load distribution characteristics; in a frontal impact, energy is mainly concentrated at connecting parts such as latches, while a side impact causes greater stress on the edges of the battery compartment.

[0061] Preferably, the difference in material impedance plays a crucial role in stress wave propagation analysis. When a stress wave propagates from one material to another, some energy is reflected back due to the difference in acoustic impedance. Impedance is defined as the product of material density and wave velocity. The impedance difference between ABS plastic and the air interface is significant, leading to strong reflection. Transmittance and reflectance are calculated based on the impedance matching principle; these coefficients determine the final strain level borne by each functional region. Normalization ensures the comparability of the influence weights of different functional regions. By dividing the comprehensive index of each region by the sum of the indices of all regions, the resulting weight value reflects the relative importance of that region in the overall structural response.

[0062] S107. Improve the edge packaging design parameters based on the quantitative indicators of deformation buffer effect and discontinuity at the opening and closing points, verify the improvement of deformation buffer effect, and obtain the strain field distribution data of the optimized shell.

[0063] Based on the wave velocity attenuation rate of the deformation buffer effect and the discontinuity level at the opening and closing points, the original design parameters of the edge packaging, including edge thickness, snap-fit ​​spacing, and sealing groove depth, are extracted. The change in buffer effect is calculated by changing a single parameter while keeping others constant. The change is divided by the parameter change to obtain the influence coefficient of each parameter. Different proportions of adjustment are applied to the three parameters according to the magnitude of the influence coefficient, resulting in an improved set of edge packaging parameter values. A new shell geometric model is constructed using the improved edge packaging parameter value set. Under the same impact load conditions, the strain response of the new shell is calculated using the finite element method. The strain peak values ​​at the snap-fit ​​positions and the battery compartment edge are extracted. The difference between the strain peak values ​​before and after the improvement is divided by the original strain peak value to obtain the percentage improvement in buffer effect. This percentage improvement value and the corresponding parameter configuration are recorded. Based on the parameter configuration with the largest percentage improvement in buffer effect, a complete finite element model of the optimized shell is established. The same impact load and constraints as the original test are applied. The displacement and strain values ​​at each time point are calculated using the time integration method. The strain data of all nodes are arranged according to spatial coordinates and time series, outputting optimized shell strain field distribution data containing complete spatiotemporal information.

[0064] In one possible implementation, the calculation of the parameter influence coefficient employs the basic principle of the controlled variable method. When assessing the impact of edge thickness on the buffering effect, the snap-fit ​​spacing and sealing groove depth are kept constant, and only a small amount is added to the edge thickness. Then, the wave velocity attenuation rate of the entire system is recalculated. The difference between the two calculations reflects the change in buffering effect caused by the change in edge thickness. Dividing this change in effect by the change in thickness yields the edge thickness influence coefficient. The advantage of this method is that it accurately identifies the independent contribution of each design parameter, avoiding evaluation bias caused by mutual interference between parameters.

[0065] Specifically, the adjustment strategy for the three design parameters is based on the relative magnitude of their influence coefficients. Assuming the edge thickness has the largest influence coefficient, it indicates that increasing the thickness is most effective in improving the buffering effect, and therefore receives the largest adjustment. The snap-fit ​​spacing may have a moderate impact, and its adjustment range is correspondingly smaller. The sealing groove depth has the smallest impact, requiring only minor adjustments. This differentiated adjustment avoids material waste and weight increases caused by blindly increasing all parameters, achieving a balance between performance and cost.

[0066] It should be noted that the application of the finite element method in calculating the strain response of the new shell involves a complex numerical solution process. The improved geometric model is divided into a large number of tiny elements, each with a stress-strain relationship that follows the material constitutive equations. By assembling the stiffness matrices of all elements, an overall equilibrium equation is formed. Under impact loading, this equation system is solved to obtain the displacements at each node, and then the strain distribution is calculated. The latching locations and the edge of the battery compartment are key monitoring points, and their strain peak values ​​directly reflect the structural safety margin.

[0067] In one embodiment, calculating the percentage improvement in buffer effect provides a quantitative assessment of the improvement. The original design might produce a maximum strain value of 0.05 under impact; after parameter optimization, this is reduced to 0.035, representing a 30% improvement. This value not only verifies the effectiveness of the improvement but also provides a benchmark for further optimization. Recording the percentage improvement corresponding to different parameter configurations allows for the establishment of a mapping library between design parameters and performance improvements.

[0068] Preferably, the time integration method plays a central role in dynamic response calculation. The remote control drop is a transient process, with strain changing rapidly over time. By discretizing time into a series of small time steps and solving the dynamic equations within each time step, the response history of the entire impact process is gradually obtained. This method can capture the propagation process of stress waves and the vibration characteristics of the structure, providing detailed data for understanding the dissipation mechanism of impact energy. The output strain field distribution data has complete spatiotemporal information characteristics. Spatially, each node has a defined three-dimensional coordinate and a corresponding strain tensor; temporally, it records the entire process from the onset of impact to vibration decay.

[0069] S108. If the strain gradient at the opening and closing points in the optimized strain field distribution data exceeds the threshold, the impact data of the improved scheme will be re-acquired, the color mapping parameters will be adjusted, the fourth visualization result will be obtained, and the result will be compared with the initial model to obtain a design improvement effect evaluation report.

[0070] The strain gradient values ​​of the opening and closing areas are extracted from the optimized strain field distribution data. The ratio of the strain difference between adjacent nodes to the distance is calculated. If this ratio exceeds a preset threshold, an impact test is conducted on the improved remote control housing. A high-speed camera re-acquires the deformation image sequence of the edge encapsulation under the impact load, obtaining new impact data containing timestamps and spatial coordinates. Based on the strain distribution characteristics in the new impact data, the starting and ending values ​​of the color mapping are adjusted, the number of color levels in high strain gradient areas is increased, and the display contrast at the opening and closing areas is enhanced through a nonlinear mapping function. The adjusted mapping parameters are applied to the entire strain field data to generate a fourth visualization result that highlights the strain distribution in key areas. The fourth visualization result is compared pixel-by-pixel with the original visualization result before optimization. The strain value difference and strain gradient improvement rate at the same location are calculated. The percentage reduction of strain peak values ​​at the buckle position, battery compartment edge, and around the button hole is statistically analyzed. All quantitative indicators are integrated to form a design improvement effect evaluation report that includes comparison data before and after improvement, strain change trends at key locations, and the overall improvement in buffering effect.

[0071] In one possible implementation, the strain gradient threshold is set based on a combination of material failure theory and engineering experience. The opening / closing point, as a geometrically discontinuous region of the remote control housing, often exhibits a higher strain gradient than other locations. When the ratio of the strain difference between adjacent nodes to their distance reaches a certain critical value, it indicates severe stress concentration in that region, potentially leading to crack initiation. This threshold is typically set to 80% of the material's yield strain gradient, ensuring a safety margin without being overly conservative and causing unnecessary redesign.

[0072] Specifically, the necessity of re-conducting impact tests lies in verifying the actual effectiveness of the improved design. Although finite element simulation can predict strain distribution, factors such as process deviations and material inhomogeneities in the actual manufacturing process may lead to differences between measured results and simulations. The actual deformation process captured by high-speed cameras can reflect these subtle but important differences, especially in complex geometric regions such as opening and closing points, where the value of measured data is even more prominent.

[0073] It's worth noting that nonlinear mapping functions have unique advantages in enhancing display contrast. Traditional linear mapping, when dealing with a large strain range, may lead to loss of detail in low-strain areas or saturation in high-strain areas. Nonlinear mapping, through logarithmic, power, or piecewise functions, can highlight details in key areas while maintaining the overall display quality.

[0074] For example, denser color gradation can be used in regions with high strain gradients so that even minute strain changes can be clearly shown through color differences.

[0075] In one embodiment, pixel-by-pixel comparative analysis provides the most refined method for evaluating the improvement effect. Each pixel represents a tiny region on the housing, and by comparing the strain values ​​at the same location before and after the improvement, the local effect of the improvement can be precisely quantified. The calculation of the strain gradient improvement rate considers not only the reduction in strain value but also the degree of uniformity in strain distribution. A successful improvement design should simultaneously achieve a reduction in peak strain and a smoothing of the strain gradient.

[0076] Preferably, the structured presentation of the design improvement effect evaluation report enables decision-makers to quickly grasp key information. Quantitative indicators in the report include the percentage reduction in peak strain at each key location, directly reflecting the degree of improvement in structural safety. The strain change trend graph shows the improvement in dynamic response throughout the entire process from the onset of impact to vibration decay. The overall buffering effect improvement evaluates the comprehensive effect of the design improvement from a macroscopic perspective. This multi-dimensional evaluation method ensures the comprehensiveness and reliability of the improvement effect. By comparing the improvements in different functional areas such as the buckle location, battery compartment edge, and around the button hole, it is possible to identify which areas benefit the most and which areas still need further optimization.

[0077] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for extracting impact deformation features based on high-speed photography and image processing, characterized in that, The method includes: The transient deformation data of the remote control housing edge encapsulated under impact is collected, the strain rate change is recorded, the material properties, encapsulation structure size and geometric discontinuity features at the opening and closing are obtained, and a continuous frame sequence image is generated. Preprocess continuous frame sequence images, extract stress wave reflection and deformation buffer effect features, identify vibration modes of thin-walled shell structure and stress concentration areas at the edge of battery compartment, and obtain the first processed image set; Based on the first processed image set, a strain field visualization model is constructed. The deformation buffer effect characteristics, stress wave reflection, shell thin wall thickness distribution, and key hole stiffness non-uniformity are input into the strain field visualization model. The shell geometric features are mapped through finite element mesh, and the strain gradient distribution is output. Based on the material properties, color mapping parameters are set, and the strain gradient distribution is judged in combination with the gradient display accuracy. The strain field visualization model is iteratively optimized to generate a second visualization result that reflects the discontinuity at the opening and closing points. Frequency domain analysis is performed on the second visualization result to obtain the stress wave reflection interference mode resolution. The local contrast of the stress concentration area and the edge discontinuity position is adjusted. The gradient display accuracy is adjusted according to the strain rate change to obtain the third visualization result. The third visualization result is subjected to strain field data quantization processing to extract strain field feature data, and the influence weight of stress wave reflection on the strain field of the shell functional area is determined by combining the drop height and impact angle. Adjust the edge packaging design parameters based on the quantitative indicators of the deformation buffer effect and the discontinuity at the opening and closing points, and obtain the optimized shell strain field distribution data; The strain gradient at the opening and closing points in the optimized strain field distribution data is judged, the impact data of the improved scheme is collected, the color mapping parameters are adjusted, and the fourth visualization result is obtained.

2. The method for extracting impact deformation features based on high-speed photography and image processing according to claim 1, characterized in that, The process involves collecting transient deformation data of the remote control housing edge encapsulated under impact, recording strain rate changes, acquiring material properties, encapsulation structure dimensions, and geometric discontinuities at the opening and closing points, and generating a continuous frame sequence of images, including: The dynamic response of the shell edge under impact is captured by high-speed camera, and the displacement time history curves of each monitoring point are obtained. The instantaneous deformation velocity is calculated based on the displacement difference and time interval between adjacent frames. The strain rate distribution data is obtained based on the ratio of the instantaneous deformation velocity to the edge size, and the continuous frame sequence image is generated. The strain rate peak region is extracted from the continuous frame sequence image, and grayscale gradient analysis is performed on the image at the corresponding time of the peak region. The edge stress distribution is calculated by tracking the pixel displacement. The plastic deformation region is determined according to the material properties, and the critical strain rate and spatial coordinates are recorded. Based on the critical strain rate and spatial coordinates, combined with the geometric discontinuity characteristics of the opening and closing point, the stress amplification factor at the buckle position is calculated to determine the risk of buckle detachment and identify the weak point.

3. The method for extracting impact deformation features based on high-speed photography and image processing according to claim 1, characterized in that, The preprocessed continuous frame sequence images are used to extract stress wave reflection and deformation buffering effect features, identify vibration modes of the thin-walled shell structure and stress concentration areas at the battery compartment edges, and obtain a first processed image set, including: The continuous frame sequence images are denoised by median filtering, and the median value of the neighborhood of each pixel is calculated to replace the center pixel value to obtain a denoised image. The second derivative value of each pixel in the denoised image is calculated by the Laplacian operator and superimposed on the corresponding pixel value to obtain an edge-enhanced image. Gray-level abrupt change regions are extracted from the edge-enhanced image, the gray-level difference between adjacent pixels is calculated, the coordinates and timestamps of the stress wave propagation path are recorded, the wave velocity is calculated based on the distance and time interval between wave crests, the deformation buffering effect is determined, and stress wave reflection feature data is obtained. The stress wave reflection feature data is transformed in the frequency domain to extract the frequency component amplitude and determine the dominant mode of the thin-walled structure. Based on the dominant mode and the buffering effect, the stress gradient is calculated, the stress concentration region at the edge of the battery compartment is marked, and the first processed image set is generated.

4. The method for extracting impact deformation features based on high-speed photography and image processing according to claim 1, characterized in that, The process involves constructing a strain field visualization model based on the first processed image set. This model inputs deformation buffering effect characteristics, stress wave reflection, shell thin-wall thickness distribution, and keyhole stiffness inhomogeneity into the strain field visualization model. By mapping the shell's geometric features through a finite element mesh, the strain gradient distribution is output, including: Pixel coordinates and grayscale values ​​of stress concentration regions are extracted from the first processed image set. A strain field data matrix is ​​constructed based on the strain-grayscale ratio. Deformation buffer effect features are superimposed to obtain the initial strain field. Boundary conditions are set according to the stress wave reflection points. The thickness distribution of the shell thin wall is calculated. The stiffness reduction coefficient of the key hole is identified to generate a structural feature dataset. Based on the structural feature dataset, the shell surface is divided by finite element mesh, the initial strain field data is mapped, the nodal strain gradient is calculated, and the strain gradient distribution is output.

5. The method for extracting impact deformation features based on high-speed photography and image processing according to claim 4, characterized in that, The process of dividing the shell surface into finite element meshes based on the structural feature dataset, mapping the initial strain field data, calculating the nodal strain gradients, and outputting the strain gradient distribution includes: Based on the structural feature dataset, the shell surface is meshed using quadrilateral elements, the mesh density is increased at the rounded corners, the mesh arrangement is adjusted along the stiffeners, and fixed constraints are set at the screw columns; the initial strain field data is mapped to the mesh nodes through shape functions, and the node strain values ​​are calculated; the node strain gradient is calculated based on the strain difference and spacing between adjacent nodes, and the strain gradient distribution is output.

6. The method for extracting impact deformation features based on high-speed photography and image processing according to claim 1, characterized in that, The process involves setting color mapping parameters based on material properties, determining strain gradient distribution by combining gradient display accuracy, iteratively optimizing the strain field visualization model, and generating a second visualization result reflecting the discontinuity at the opening and closing points. This includes: The color mapping range is set according to the material properties, the strain gradient distribution is mapped to the RGB color space, the RGB value distance between mesh cells is calculated, the contrast is adjusted, and a preliminary visualization image is obtained. The geometric boundary at the opening and closing point is identified from the preliminary visualization image, the discontinuity is measured, displacement constraints are added, and local strain field correction data is calculated. The mesh node density and shape function are adjusted, the color mapping is iteratively optimized, and the second visualization result is generated.

7. The method for extracting impact deformation features based on high-speed photography and image processing according to claim 6, characterized in that, The process of adjusting the grid node density and shape function, iteratively optimizing the color mapping, and generating a second visualization result includes: Based on the local strain field correction data, the grid node density at the opening and closing points is increased, the shape function order is improved, and the correction data is mapped to the visualization image; the nonlinear parameters of the color mapping function are adjusted, the pixel difference rate between adjacent iterative images is calculated, and the iteration is stopped; the optimized color mapping is applied to the strain field to generate the second visualization result.

8. The method for extracting impact deformation features based on high-speed photography and image processing according to claim 1, characterized in that, The process of adjusting the edge packaging design parameters based on the deformation buffer effect and the quantitative indicators of discontinuities at the opening and closing points, and obtaining optimized shell strain field distribution data, includes: Based on the deformation buffer effect and the discontinuity at the opening and closing points, edge encapsulation parameters are extracted, and the parameter values ​​are adjusted to obtain an improved parameter set. A shell geometric model is constructed based on the improved parameter set, the strain response is calculated, the strain peak value is extracted, and the percentage increase in buffer effect is calculated. Based on the parameter with the largest percentage increase, the optimized shell strain field distribution data is calculated.

9. The method for extracting impact deformation features based on high-speed photography and image processing according to claim 8, characterized in that, The step of calculating the optimized shell strain field distribution data based on the parameter with the largest percentage increase includes: Based on the parameter with the largest increase percentage, an optimized shell finite element model is established, and impact loads and constraints are applied; nodal strain values ​​are calculated by time integration; the nodal strain values ​​are arranged according to spatial coordinates and time series, and the optimized shell strain field distribution data is output.

Images