CN115859405B - Self-supporting steel chimney design data enhancement method - Google Patents
Self-supporting steel chimney design data enhancement method Download PDFInfo
- Publication number
- CN115859405B CN115859405B CN202310186893.2A CN202310186893A CN115859405B CN 115859405 B CN115859405 B CN 115859405B CN 202310186893 A CN202310186893 A CN 202310186893A CN 115859405 B CN115859405 B CN 115859405B
- Authority
- CN
- China
- Prior art keywords
- curve
- point
- interpolation
- points
- hilbert curve
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 229910000831 Steel Inorganic materials 0.000 title claims abstract description 67
- 239000010959 steel Substances 0.000 title claims abstract description 67
- 238000000034 method Methods 0.000 title claims abstract description 60
- 238000013461 design Methods 0.000 title claims abstract description 20
- 230000002159 abnormal effect Effects 0.000 claims abstract description 106
- 238000012545 processing Methods 0.000 claims abstract description 28
- 230000002708 enhancing effect Effects 0.000 claims abstract description 12
- 230000000694 effects Effects 0.000 abstract description 28
- 230000007704 transition Effects 0.000 description 20
- 238000005070 sampling Methods 0.000 description 7
- 230000003044 adaptive effect Effects 0.000 description 5
- 238000011156 evaluation Methods 0.000 description 3
- 238000011176 pooling Methods 0.000 description 3
- 238000012935 Averaging Methods 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
Images
Landscapes
- Image Processing (AREA)
Abstract
The invention relates to the technical field of image processing, in particular to a method for enhancing design data of a self-supporting steel chimney. The method comprises the following steps: acquiring an initial image of a self-standing steel chimney area, and downsampling the initial image to acquire a downsampled image, and acquiring a first Hilbert curve corresponding to the initial image and a second interpolation curve corresponding to the downsampled image; respectively matching each second interpolation curve with the first Hilbert curve to obtain abnormal points on each second interpolation curve, constructing a characteristic point set corresponding to each abnormal point, and further obtaining target values of each abnormal point; and determining weights of adjacent points on the left and right sides of the corresponding position of each abnormal point on the first Hilbert curve according to the target value and curve values of adjacent points on the left and right sides of the corresponding position of each abnormal point on the first Hilbert curve, so as to obtain an enhanced image. The invention improves the enhancement effect of the image data of the self-standing steel chimney.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to a method for enhancing design data of a self-supporting steel chimney.
Background
After a CAD drawing is used for drawing a design drawing of the self-supporting steel chimney, the CAD drawing is required to be output as a high-resolution grating image, when the resolution of the output image cannot meet the requirement, image processing is usually carried out by means of post-processing software such as Photoshop and the like, the image resolution is improved by adopting bilinear interpolation in a conventional method, the interpolation is a new data derived on the basis of original data, so that the resolution is improved, but when bilinear interpolation is adopted for four adjacent pixel points, the obtained surfaces are consistent at a neighborhood, but the slopes are not consistent, and the detail of the image, namely the natural transitional effect, is degraded due to the smoothing effect of the bilinear gray interpolation, and the phenomenon is particularly obvious when the image is amplified, so that the enhancement effect of the existing bilinear gray interpolation method on the image data of the self-supporting steel chimney is poor.
Disclosure of Invention
In order to solve the problem that the enhancement effect is poor when the existing bilinear gray interpolation method is used for enhancing the image data of the self-supporting steel chimney, the invention aims to provide the self-supporting steel chimney design data enhancement method, and the adopted technical scheme is as follows:
the invention provides a method for enhancing design data of a self-supporting steel chimney, which comprises the following steps:
acquiring an initial image of a self-standing steel chimney area;
obtaining the size of the primitive based on the minimum circumscribed rectangle of the initial image; performing downsampling processing on the initial image for multiple times to obtain downsampled images, and obtaining a first Hilbert curve corresponding to the initial image and a second Hilbert curve corresponding to each downsampled image based on the primitive;
interpolation processing is carried out on the second Hilbert curve corresponding to each downsampled image to obtain a second interpolation curve corresponding to each downsampled image; respectively matching each second interpolation curve with the first Hilbert curve to obtain abnormal points on each second interpolation curve; taking a set formed by matching points of each abnormal point as a characteristic point set corresponding to each abnormal point, and obtaining target values of each abnormal point based on the characteristic point set and curve values of adjacent points on the left side and the right side of a corresponding position of each abnormal point on the first Hilbert curve; according to the target value and curve values of adjacent points on the left side and the right side of the corresponding position of each abnormal point on the first Hilbert curve, determining weights of the adjacent points on the left side and the right side of the corresponding position of each abnormal point on the first Hilbert curve;
and carrying out interpolation processing on the initial image based on the weight to obtain an enhanced image of the free-standing steel chimney area.
Preferably, the obtaining the size of the primitive based on the minimum bounding rectangle of the initial image includes:
calculating the aspect ratio of the minimum circumscribed rectangle of the initial image; the numerator in the aspect ratio is taken as the length of the primitive, and the denominator in the aspect ratio is taken as the width of the primitive.
Preferably, the matching the second interpolation curves with the first hilbert curves to obtain abnormal points on the second interpolation curves includes:
acquiring a matching relationship between a point on a second interpolation curve and a point on the first Hilbert curve;
for any point on any of the second interpolation curves: and judging whether the number of the points matched with the points on the first Hilbert curve is larger than 1, and if so, taking the points as abnormal points on the second interpolation curve.
Preferably, the obtaining the target value of each abnormal point based on the characteristic point set and curve values of adjacent points on the left and right sides of the corresponding position of each abnormal point on the first hilbert curve includes:
a pixel point at the lower left corner of the first Hilbert curve is taken as a coordinate origin, the horizontal direction passing through the coordinate origin is taken as an X axis of a coordinate system, and the direction passing through the coordinate origin and perpendicular to the X axis is taken as a Y axis of the coordinate system, so that a rectangular coordinate system is constructed;
for the ith outlier on the second interpolation curve:
taking the coordinates of all points in the feature point set corresponding to the ith abnormal point as the input of PCA, obtaining the principal component direction of the feature point in the feature point set by using a PCA algorithm, marking the principal component direction with the largest feature value as the largest principal direction, and marking the slope corresponding to the largest principal direction as the target slope corresponding to the ith abnormal point;
making a straight line with a slope as a target slope on the first Hilbert curve, wherein the distance between two adjacent points on the left side and the right side of the corresponding position of the ith abnormal point on the first Hilbert curve is equal, and marking the straight line as a first straight line corresponding to the abnormal point; making a straight line perpendicular to the first straight line at the corresponding position of the ith abnormal point on the first Hilbert curve, and marking the straight line as a second straight line corresponding to the ith abnormal point; the curve value of the intersection point of the first straight line and the second straight line is taken as the target value of the ith abnormal point.
Preferably, the determining weights of the adjacent points on the left and right sides of the corresponding position of each abnormal point on the first hilbert curve according to the target value and curve values of the adjacent points on the left and right sides of the corresponding position of each abnormal point on the first hilbert curve includes:
for the ith outlier on the second interpolation curve:
according to the target value of the ith abnormal point and the values of the adjacent points on the left side and the right side of the corresponding position of the ith abnormal point on the first Hilbert curve, the following target equation is constructed:
wherein ,
for the curve value of the adjacent point to the left of the corresponding position of the ith outlier on the first hilbert curve,
for the curve value of the adjacent point to the right of the corresponding position on the first hilbert curve for the i-th outlier,
the weight of the adjacent point to the left of the corresponding position on the first hilbert curve for the i-th outlier,
for the weight of the neighbor to the right of the corresponding position on the first hilbert curve for the i-th outlier,
a target value for the ith outlier on the second interpolation curve;
and obtaining weights of adjacent points on the left side and the right side of the corresponding position of the ith abnormal point on the first Hilbert curve based on the target equation.
Preferably, the interpolation processing is performed on the initial image based on the weight, and an enhanced image of the free-standing steel chimney area is obtained, which comprises the following steps:
for any point on the first hilbert curve: if the number of the weights of the points is larger than 1, taking the average value of the weights of the points as the target weight of the points; if the number of the weights of the points is equal to 1, the weights of the points are used as target weights of the points;
and carrying out interpolation processing on the initial image by adopting a bilinear interpolation method based on the target weights of each point on the first Hilbert curve, and recording the image after interpolation as an enhanced image of the self-standing steel chimney area.
Preferably, the length of the second interpolation curve is the same as the length of the first hilbert curve.
The invention has at least the following beneficial effects:
the primitive is built based on the minimum circumscribed rectangle of the initial image of the self-supporting steel chimney area, so that the chimney area in the minimum circumscribed rectangle can be used for building the Hilbert curve, and therefore a first Hilbert curve corresponding to the initial image of the self-supporting steel chimney area is built based on the primitive; the invention aims to improve the resolution of an initial image of a self-supporting steel chimney area, in order to achieve the aim, the embodiment carries out interpolation processing on the initial image, and takes the fact that the inverse process of downsampling is an interpolation process into consideration, so that the invention carries out downsampling processing on the initial image of the self-supporting steel chimney area to obtain a plurality of Zhang Xia sampled images, and builds a second Hilbert curve corresponding to each downsampled image; then, each second interpolation curve is matched with the first Hilbert curve, an abnormal point is obtained based on a matching result, and the natural transition effect of the position of the abnormal point is abnormal, so that in order to ensure the subsequent interpolation effect, the abnormal point and the information of the adjacent points of the abnormal point are combined to obtain the corresponding weight in interpolation processing; when interpolation is carried out on an initial image, when the difference value between a predicted value and an actual value is minimum, the prediction effect is optimal, but the interpolation is not actually compared with the actual value, so that the weight of adjacent pixel points of each position for interpolation is calculated based on the change of Hilbert curves of different downsampling layers, the transition effect is better, and the matching result of a second interpolation curve corresponding to a downsampled image and a first Hilbert curve can be used as a reference of the actual interpolation to regulate and control the actual interpolation process because the downsampled image and the original image are closely related; the method obtains the target value of the abnormal point based on the abnormal point and the matching point of the abnormal point, further determines the weights of the adjacent points on the left side and the right side of the abnormal point, interpolates the initial image based on the weights, adjusts the interpolation process by using the Hilbert curve as an evaluation index for whether transition is stable or not, further endows different interpolation weights for different points, ensures better transition effect after interpolation, and improves the enhancement effect of the image data of the self-supporting steel chimney.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a method for enhancing design data of a free-standing steel chimney according to an embodiment of the present invention;
fig. 2 is a schematic diagram of a first straight line corresponding to the ith abnormal point.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following describes a method for enhancing design data of a self-standing steel chimney according to the invention in detail with reference to the attached drawings and the preferred embodiment.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following specifically describes a specific scheme of the method for enhancing design data of the self-standing steel chimney provided by the invention with reference to the accompanying drawings.
A free-standing steel chimney design data enhancement method embodiment:
the embodiment provides a method for enhancing design data of a self-standing steel chimney, as shown in fig. 1, the method for enhancing design data of the self-standing steel chimney comprises the following steps:
step S1, acquiring an initial image of the free-standing steel chimney area.
After a design drawing of the self-supporting steel chimney is drawn by CAD, the CAD converts analog quantity into digital quantity, and when DWG drawing files are output, the resolution is low, so that the resolution is required to be improved by interpolation. The method comprises the steps of firstly obtaining DWG image data output by CAD, constructing adaptive primitives based on the DWG image, obtaining corresponding first Hilbert curves based on the constructed adaptive primitives, then carrying out downsampling on an obtained initial image, and obtaining second Hilbert curves of each downsampled image through the adaptive primitives; interpolation is carried out on each second Hilbert curve to obtain a plurality of second interpolation curves, DTW matching is carried out on the first Hilbert curves and each second interpolation curve respectively, abnormal points are obtained based on the matching results, weights of adjacent points of each abnormal point are determined, bilinear interpolation is further carried out, interpolation images are obtained, and enhancement processing of self-supporting steel chimney design data is completed.
According to the embodiment, a design drawing of the free-standing steel chimney is drawn by utilizing the CAD, DWG image data output by the CAD is obtained, the existence of the background pixel points not only increases the calculated amount, but also can influence the enhancement effect of the subsequent chimney area considering that the DWG image data contains partial background pixel points, and the pixel point with the gray value of 255 in the DWG image is the pixel point of the background area, so that the maximum connected domain formed by the pixel points with the gray value of not 255 in the DWG image is recorded as the initial image of the free-standing steel chimney, the resolution of the initial image of the free-standing steel chimney is lower, and the embodiment is processed by an interpolation method in order to improve the resolution of the image.
Step S2, obtaining the size of the primitive based on the minimum circumscribed rectangle of the initial image; and performing downsampling processing on the initial image for multiple times to obtain downsampled images, and obtaining a first Hilbert curve corresponding to the initial image and a second Hilbert curve corresponding to each downsampled image based on the primitive.
The chimney has a shape similar to a linear shape on a plane, so that the length-width ratio is larger, the problem of unstable transition area caused by a conventional bilinear interpolation method occurs, the difference of transition unstable information in the length-width direction after interpolation is larger, and the impression of an image after interpolation is greatly different from that of an original image, namely the defect of bilinear difference is amplified.
The Hilbert curve is a curve which represents multiple dimensions as one dimension, taking a two-dimensional plane as an example, the Hilbert curve can ensure that pixel points adjacent to the two-dimensional plane are adjacent when being pulled into one dimension, the reduced-dimension image has a natural transition effect, the chimney presents a shape which is similar to a linear shape and has a large length-width ratio, the Hilbert curve is used as an evaluation index for whether the transition is stable to adjust the interpolation process, and the good transition effect after interpolation can be ensured; when the difference of the Hilbert curves before and after interpolation is smaller, the natural transition effect of the interpolated image is better.
Because interpolation is actually a prediction, a method for predicting a pixel value at a certain position between existing pixel points based on the pixel value of the existing pixel points has the best prediction effect when the difference between the predicted value and an actual value is minimum, but the interpolation is not actually compared with the predicted value, so that the contribution degree, namely the weight, of adjacent pixel points when interpolation is performed at each position is calculated based on the change of Hilbert curves of different downsampling layers, the transition effect is better, and the result can be used as a reference of actual interpolation because the downsampled image and the original image are closely related, and further the actual interpolation process is regulated and controlled.
The hilbert curve is a curve which is full of space and obtained by traversing all points in a unit square, and the free-standing steel chimney is a rectangle with a larger length-width ratio in a planar image and has a larger phase difference with the square, so that an adaptive primitive is firstly required to be constructed, and then the corresponding hilbert curve is obtained.
Firstly, obtaining the minimum circumscribed rectangle of an initial image of a self-supporting steel chimney area, calculating the length-width ratio k of the minimum circumscribed rectangle, then taking the numerator of k as the length of a primitive, taking the denominator of k as the width of the primitive, namely obtaining the base size, wherein the primitive is the minimum element on the image, and under normal conditions, the primitive is a square, but the Hilbert curve can only calculate the equal number of ranks, so that the corresponding Hilbert curve can be obtained by setting an adaptive primitive, and then constructing the Hilbert curve corresponding to the initial image of the self-supporting steel chimney area, and marking the Hilbert curve as the first Hilbert curve corresponding to the initial image of the self-supporting steel chimney. The method of constructing the hilbert curve is the prior art, and will not be described in detail here.
The present embodiment obtains the size of the primitive in the above steps, and then subjects the primitive to
The method comprises the steps of taking an initial image of a free-standing steel chimney as a sampling window, sliding the initial image of the free-standing steel chimney by using the sampling window as a basic image, processing by adopting a mean value pooling method based on pixel points in the window to obtain an image after first mean value pooling, recording the image as a first downsampling image, respectively calculating the entropy value of the initial image of a free-standing steel chimney area and the entropy value of the first downsampling image, judging whether the entropy value of the first downsampling image is smaller than 80% of the entropy value of the initial image of the free-standing steel chimney area, and stopping if the entropy value of the first downsampling image is smaller than 80% of the entropy value of the initial image of the free-standing steel chimney areaDownsampling; if the entropy value of the first downsampled image is larger than or equal to the initial value of the self-standing steel chimney area, continuing to take the first downsampled image as a basic image, carrying out average pooling treatment on the first downsampled image by adopting the method to obtain a second downsampled image, calculating the entropy value of the second downsampled image, judging whether the entropy value of the second downsampled image is smaller than 80% of the entropy value of the initial image of the self-standing steel chimney area, and the like, stopping until the entropy value of the downsampled image is smaller than 80% of the entropy value of the initial image of the self-standing steel chimney area or the total downsampled times are equal to the preset times, and obtaining a plurality of downsampled images; it should be noted that, the preset number of times is set in this embodiment to prevent the situation that the entropy value of the downsampled image is always greater than or equal to 80% of the entropy value of the initial image of the free-standing steel chimney area, the preset number of times in this embodiment is 20 times, and in specific applications, an implementer can set according to specific situations. The averaging is equivalent to the fact that the contribution degree of each pixel point to the sampled pixel points is the same, so that the situation that the calculation amount is large due to the fact that the sampling contribution degrees are different in the follow-up process is avoided; the embodiment is provided with
The value of (2) is set to 2, and the sliding step length of the sampling window is set to 2, and in specific application, an implementer can set according to specific situations.
In this embodiment, a plurality of downsampled images are obtained, for each downsampled image, each primitive corresponds to a pixel value, and a hilbert curve corresponding to each downsampled image is obtained according to the pixel value corresponding to each primitive, and is recorded as a second hilbert curve corresponding to each downsampled image.
To this end, the present embodiment obtains one first downsampled image and a plurality of second downsampled images.
Step S3, interpolation processing is carried out on the second Hilbert curves corresponding to the downsampled images to obtain second interpolation curves corresponding to the downsampled images; respectively matching each second interpolation curve with the first Hilbert curve to obtain abnormal points on each second interpolation curve; taking a set formed by matching points of each abnormal point as a characteristic point set corresponding to each abnormal point, and obtaining target values of each abnormal point based on the characteristic point set and curve values of adjacent points on the left side and the right side of a corresponding position of each abnormal point on the first Hilbert curve; and determining weights of adjacent points on the left and right sides of the corresponding position of each abnormal point on the first Hilbert curve according to the target value and curve values of adjacent points on the left and right sides of the corresponding position of each abnormal point on the first Hilbert curve.
In the embodiment, the natural transition effect is evaluated through the change condition of the hilbert curves corresponding to the images before and after sampling, and then the weight when interpolation is carried out on different pixel points is calculated, but because the lengths of the hilbert curves are different due to the change of the number of the pixel points in the images before and after sampling, the interpolation is carried out on each second hilbert curve by adopting the existing interpolation method, so that the length of each second hilbert curve after interpolation processing is the same as the length of the first hilbert curve, each second hilbert curve after interpolation is marked as a second interpolation curve, and a second interpolation image corresponding to each downsampled image is obtained; for example: if the length of the first hilbert curve is 1000, the length of a certain second hilbert curve is 250, and the nearest interpolation is performed on the second hilbert curve by adopting the existing interpolation method, so that the length of the second hilbert curve after the interpolation processing is also 1000. In this embodiment, the first hilbert curve is not sampled, but the second hilbert curve is interpolated, because the first hilbert curve carries standard transition information, the information may be destroyed after sampling, and when the second hilbert curve is compared with the sampled first hilbert curve, a great difference exists between the comparison result and the actual result, so that the subsequent enhancement effect is affected.
In the embodiment, a second interpolation curve is obtained for each second hilbert curve through nearest neighbor interpolation, then the DTW distance between the first hilbert curve and each second interpolation curve is calculated, and in the process of calculating the DTW distance between the first hilbert curve and each second interpolation curve, a corresponding relation can be obtained; and acquiring points, corresponding to a plurality of points on the first Hilbert curve, of one point on the second interpolation curve in the DTW matching, as abnormal points of the second interpolation curve, namely, taking the point of one on the second interpolation curve in the one-to-many process of the second interpolation curve in the matching process as the abnormal point of the second interpolation curve, wherein the transition effect of the one-to-many area is different from that of the original image. For any point on any of the second interpolation curves: and judging whether the number of the points matched with the points on the first Hilbert curve is larger than 1, and if so, taking the points as abnormal points on the second interpolation curve. By adopting the method, the abnormal point on each second interpolation curve can be obtained.
In order to ensure that the trend of the corrected abnormal points is the same as the trend of the corresponding position of the first Hilbert curve, namely, the second interpolation curve is parallel to the first Hilbert curve, so that the requirement of consistent with the original transition stability is met.
For the ith outlier on the second interpolation curve: the method comprises the steps that an ith abnormal point corresponds to a plurality of matching points on a first Hilbert curve, the matching points of the ith abnormal point on the first Hilbert curve are marked as characteristic points, all the matching points of the ith abnormal point on the first Hilbert curve are used as a set, the set of the characteristic points corresponding to the ith abnormal point is marked, a pixel point at the lower left corner of the first Hilbert curve is used as a coordinate origin, the horizontal positive direction passing through the coordinate origin is used as an X axis of a coordinate system, and the direction passing through the coordinate origin and perpendicular to the X axis is used as a Y axis of the coordinate system, so that a rectangular coordinate system is constructed; the method comprises the steps of taking coordinates of all points in a feature point set corresponding to an ith abnormal point as input of PCA, obtaining principal component directions of the feature points in the feature point set by using a PCA algorithm, obtaining a plurality of principal component directions, wherein each principal component direction is a 2-dimensional unit vector, each principal component direction corresponds to a feature value, obtaining the principal component direction with the largest feature value and marking the principal component direction as the largest principal direction, representing the main distribution direction of all the feature points in the feature point set, obtaining a slope corresponding to the largest principal direction, marking the slope as a target slope corresponding to the ith abnormal point, making a slope on a first Hilbert curve as a target slope, marking a straight line with equal distance between two adjacent points on the left side and the right side of the corresponding position of the ith abnormal point on the first Hilbert curve as a first straight line corresponding to the abnormal point, and marking the straight line as a straight line L in fig. 2; making a straight line perpendicular to the first straight line at the corresponding position of the ith abnormal point on the first Hilbert curve, and marking the straight line as a second straight line corresponding to the ith abnormal point; taking a curve value of an intersection point of the first straight line and the second straight line as a target value of an ith abnormal point; next, in this embodiment, the weight of the adjacent points on the left and right sides of the corresponding position of each abnormal point on the first hilbert curve is determined based on the target value of the i-th abnormal point and the values of the two adjacent points of the i-th abnormal point on the first hilbert curve, as shown in fig. 2, the point i in the graph represents the i-th abnormal point, the point i-1 in the graph represents the point adjacent to the left side of the corresponding position of the i-th abnormal point on the first hilbert curve, and the point i+1 in the graph represents the point adjacent to the right side of the corresponding position of the i-th abnormal point on the first hilbert curve; the resolution of the initial image of the free-standing steel chimney is low, and in this embodiment, in order to improve the resolution of the initial image of the free-standing steel chimney region, the downsampling process is considered to be an interpolation process, so that the actual interpolation process is provided with a basis through the downsampling process, that is, the resolution of the image is increased through interpolation of the initial image of the free-standing steel chimney region. Based on this, in this embodiment, according to the target value of the i-th outlier and the values of the adjacent points on the left and right sides of the corresponding position of the i-th outlier on the first hilbert curve, a target equation is constructed, where the target equation is specifically:
wherein ,
the curve value of the adjacent point on the left side of the corresponding position of the ith abnormal point on the first Hilbert curve, namely the ordinate of the adjacent point on the left side of the corresponding position of the ith abnormal point on the first Hilbert curve,
for the curve value of the adjacent point to the right of the corresponding position of the i-th outlier on the first Hilbert curve, that is to say the ordinate of the adjacent point to the right of the corresponding position of the i-th outlier on the first Hilbert curve,
the weight of the adjacent point to the left of the corresponding position on the first hilbert curve for the i-th outlier,
for the weight of the neighbor to the right of the corresponding position on the first hilbert curve for the i-th outlier,
a target value for the ith outlier on the second interpolation curve; according to the objective equation constructed in the embodiment, weights of adjacent points on the left and right sides of the corresponding position of the ith abnormal point on the first Hilbert curve can be calculated.
By adopting the method, the weights of two adjacent points on the left side and the right side of the corresponding position of each abnormal point on each second interpolation curve on the first Hilbert curve can be obtained.
And S4, carrying out interpolation processing on the initial image based on the weight to obtain an enhanced image of the free-standing steel chimney area.
Since the number of the second interpolation curves is more than one, a plurality of weights may exist at a part of points on the first Hilbert curve, and when a certain point exists a plurality of weights, the average value of the weights of the point is taken as the target weight of the point; when a certain point has a weight, directly taking the weight of the point as the target weight of the point; it should be noted that, the points at the upper part of the first hilbert curve are not adjacent to the abnormal points, which means that the natural transition effect of the areas where the points are located is normal and no abnormality occurs, and for such points, the method provided by the embodiment cannot acquire the target weight, and then when bilinear interpolation is performed, the original weight is directly used as the weight when bilinear interpolation is performed to perform interpolation processing; since the point on the first hilbert curve has a correspondence relationship with the pixel point in the initial image of the free-standing steel chimney region, the target weight of the pixel point in the initial image of the free-standing steel chimney region is obtained.
The target weight of each point on the first Hilbert-Tech curve is used as the initial weight of the corresponding pixel point in the initial image of the self-standing steel chimney area during bilinear interpolation, and then bilinear interpolation is carried out on the basis of the initial weight to obtain an interpolation result, the image after interpolation is recorded as an enhanced image of the self-standing steel chimney, and the bilinear interpolation is the prior art, and only the initial weight during bilinear interpolation is newly given, so that redundant description is omitted. The method provided by the embodiment completes the enhancement processing of the initial image of the self-supporting steel chimney, and improves the resolution of design data of the self-supporting steel chimney.
In the embodiment, the primitive is constructed based on the minimum circumscribed rectangle of the initial image of the self-supporting steel chimney area, so that the chimney area in the minimum circumscribed rectangle can be constructed by the Hilbert curve, and therefore, the first Hilbert curve corresponding to the initial image of the self-supporting steel chimney area is constructed based on the primitive; the purpose of this embodiment is to improve the resolution of the initial image of the self-supporting steel chimney area, in order to achieve this purpose, this embodiment will perform interpolation processing on the initial image, considering that the inverse process of downsampling is an interpolation process, so this embodiment performs downsampling processing on the initial image of the self-supporting steel chimney area to obtain a plurality of Zhang Xia sampled images, and builds a second hilbert curve corresponding to each downsampled image, since the first hilbert curve is built based on the initial image, and the second hilbert curve is built based on the downsampled image, the length of the second hilbert curve is smaller than the length of the first hilbert curve, and then matches the first hilbert curve with the second hilbert curve to obtain an outlier, so the existing interpolation method adopted in this embodiment performs interpolation processing on each second hilbert curve to obtain a second interpolation curve corresponding to each downsampled image; then, each second interpolation curve is matched with the first Hilbert curve, an abnormal point is obtained based on a matching result, and the natural transition effect of the position of the abnormal point is abnormal, so that in order to ensure the subsequent interpolation effect, the abnormal point and the information of the adjacent points of the abnormal point are combined to obtain the corresponding weight in interpolation processing; when interpolation is carried out on an initial image, when the difference value between a predicted value and an actual value is minimum, the prediction effect is optimal, but the interpolation is not actually compared with the actual value, so that the weight of adjacent pixel points of each position for interpolation is calculated based on the change of Hilbert curves of different downsampling layers, the transition effect is better, and the matching result of a second interpolation curve corresponding to a downsampled image and a first Hilbert curve can be used as a reference of the actual interpolation to regulate and control the actual interpolation process because the downsampled image and the original image are closely related; according to the embodiment, the target value of the abnormal point is obtained based on the abnormal point and the matching point of the abnormal point, the weight of the adjacent points on the left side and the right side of the abnormal point is further determined, the initial image is interpolated based on the weight, the Hilbert curve is used as an evaluation index for whether transition is stable or not to adjust the interpolation process, and further different interpolation weights of different points are given, so that the effect of the interpolated transition is better, and the enhancement effect of the self-supporting steel chimney image data is improved.
Claims (6)
1. A method of enhancing design data of a free-standing steel stack, the method comprising the steps of:
acquiring an initial image of a self-standing steel chimney area;
obtaining the size of the primitive based on the minimum circumscribed rectangle of the initial image; performing downsampling processing on the initial image for multiple times to obtain downsampled images, and obtaining a first Hilbert curve corresponding to the initial image and a second Hilbert curve corresponding to each downsampled image based on the primitive;
interpolation processing is carried out on the second Hilbert curve corresponding to each downsampled image to obtain a second interpolation curve corresponding to each downsampled image; respectively matching each second interpolation curve with the first Hilbert curve to obtain abnormal points on each second interpolation curve; taking a set formed by matching points of each abnormal point as a characteristic point set corresponding to each abnormal point, and obtaining target values of each abnormal point based on the characteristic point set and curve values of adjacent points on the left side and the right side of a corresponding position of each abnormal point on the first Hilbert curve; according to the target value and curve values of adjacent points on the left side and the right side of the corresponding position of each abnormal point on the first Hilbert curve, determining weights of the adjacent points on the left side and the right side of the corresponding position of each abnormal point on the first Hilbert curve;
interpolation processing is carried out on the initial image based on the weight, and an enhanced image of the free-standing steel chimney area is obtained;
the obtaining the target value of each abnormal point based on the characteristic point set and the curve values of the adjacent points on the left side and the right side of the corresponding position of each abnormal point on the first Hilbert curve comprises the following steps:
a pixel point at the lower left corner of the first Hilbert curve is taken as a coordinate origin, the horizontal direction passing through the coordinate origin is taken as an X axis of a coordinate system, and the direction passing through the coordinate origin and perpendicular to the X axis is taken as a Y axis of the coordinate system, so that a rectangular coordinate system is constructed;
for the ith outlier on the second interpolation curve:
taking the coordinates of all points in the feature point set corresponding to the ith abnormal point as the input of PCA, obtaining the principal component direction of the feature point in the feature point set by using a PCA algorithm, marking the principal component direction with the largest feature value as the largest principal direction, and marking the slope corresponding to the largest principal direction as the target slope corresponding to the ith abnormal point;
making a straight line with a slope as a target slope on the first Hilbert curve, wherein the distance between two adjacent points on the left side and the right side of the corresponding position of the ith abnormal point on the first Hilbert curve is equal, and marking the straight line as a first straight line corresponding to the abnormal point; making a straight line perpendicular to the first straight line at the corresponding position of the ith abnormal point on the first Hilbert curve, and marking the straight line as a second straight line corresponding to the ith abnormal point; the curve value of the intersection point of the first straight line and the second straight line is taken as the target value of the ith abnormal point.
2. The method of claim 1, wherein obtaining the dimensions of the primitive based on the minimum bounding rectangle of the initial image comprises:
calculating the aspect ratio of the minimum circumscribed rectangle of the initial image; the numerator in the aspect ratio is taken as the length of the primitive, and the denominator in the aspect ratio is taken as the width of the primitive.
3. The method for enhancing design data of a free-standing steel chimney according to claim 1, wherein the matching each second interpolation curve with the first hilbert curve to obtain an abnormal point on each second interpolation curve comprises:
acquiring a matching relationship between a point on a second interpolation curve and a point on the first Hilbert curve;
for any point on any of the second interpolation curves: and judging whether the number of the points matched with the points on the first Hilbert curve is larger than 1, and if so, taking the points as abnormal points on the second interpolation curve.
4. The method for enhancing design data of a free-standing steel chimney according to claim 1, wherein determining weights of adjacent points on the left and right sides of the corresponding position of each abnormal point on the first hilbert curve according to the target value and curve values of adjacent points on the left and right sides of the corresponding position of each abnormal point on the first hilbert curve comprises:
for the ith outlier on the second interpolation curve:
according to the target value of the ith abnormal point and the values of the adjacent points on the left side and the right side of the corresponding position of the ith abnormal point on the first Hilbert curve, the following target equation is constructed:
wherein ,
for the curve value of the adjacent point to the left of the corresponding position of the ith outlier on the first hilbert curve,
for the curve value of the adjacent point on the right of the corresponding position of the ith outlier on the first Hilbert curve,/for the curve value of the i-th outlier>
Weight of adjacent point on left side of corresponding position on the first Hilbert curve for ith outlier, ++>
Weight of the right neighbor point of the corresponding position on the first Hilbert curve for the ith outlier, +.>
A target value for the ith outlier on the second interpolation curve;
and obtaining weights of adjacent points on the left side and the right side of the corresponding position of the ith abnormal point on the first Hilbert curve based on the target equation.
5. The method of claim 4, wherein interpolating the initial image based on the weights to obtain an enhanced image of the free-standing steel stack area, comprising:
for any point on the first hilbert curve: if the number of the weights of the points is larger than 1, taking the average value of the weights of the points as the target weight of the points; if the number of the weights of the points is equal to 1, the weights of the points are used as target weights of the points;
and carrying out interpolation processing on the initial image by adopting a bilinear interpolation method based on the target weights of each point on the first Hilbert curve, and recording the image after interpolation as an enhanced image of the self-standing steel chimney area.
6. The method of claim 1, wherein the second interpolation curve has a length equal to a length of the first hilbert curve.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310186893.2A CN115859405B (en) | 2023-03-02 | 2023-03-02 | Self-supporting steel chimney design data enhancement method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310186893.2A CN115859405B (en) | 2023-03-02 | 2023-03-02 | Self-supporting steel chimney design data enhancement method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115859405A CN115859405A (en) | 2023-03-28 |
| CN115859405B true CN115859405B (en) | 2023-05-12 |
Family
ID=85659589
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310186893.2A Active CN115859405B (en) | 2023-03-02 | 2023-03-02 | Self-supporting steel chimney design data enhancement method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115859405B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117436283B (en) * | 2023-12-13 | 2024-03-26 | 青岛昊宇重工有限公司 | Computer aided design method, system and storage medium for chimney |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106875335A (en) * | 2017-02-16 | 2017-06-20 | 南昌航空大学 | It is a kind of improved based on Hilbert-Huang transform image background suppressing method |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8340415B2 (en) * | 2010-04-05 | 2012-12-25 | Microsoft Corporation | Generation of multi-resolution image pyramids |
| CN103218784B (en) * | 2013-04-19 | 2016-06-01 | 北京理工大学 | Based on the rapid image filtering method of space filling curve and extreme point |
| US10372728B2 (en) * | 2016-09-16 | 2019-08-06 | Oracle International Corporation | System and method providing a scalable and efficient space filling curve approach to point cloud feature generation |
| CN106447740B (en) * | 2016-10-08 | 2019-04-02 | 重庆大学 | Opposite parallel lines CT area-of-interest image rebuilding method |
| US10776982B2 (en) * | 2017-07-03 | 2020-09-15 | Artomatix Ltd. | Systems and methods for providing non-parametric texture synthesis of arbitrary shape and/or material data in a unified framework |
| CN115115615B (en) * | 2022-07-26 | 2022-12-13 | 南通好心情家用纺织品有限公司 | Textile fabric quality evaluation method and system based on image recognition |
| CN115423688A (en) * | 2022-09-29 | 2022-12-02 | 上海海事大学 | Quantum circuit diagram and quantum color image scaling method based on bilinear interpolation |
-
2023
- 2023-03-02 CN CN202310186893.2A patent/CN115859405B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106875335A (en) * | 2017-02-16 | 2017-06-20 | 南昌航空大学 | It is a kind of improved based on Hilbert-Huang transform image background suppressing method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115859405A (en) | 2023-03-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115859405B (en) | Self-supporting steel chimney design data enhancement method | |
| CN109886279B (en) | Image processing method, device, computer equipment and storage medium | |
| CN102521794B (en) | Image interpolation method and device based on spline surface | |
| CN104700360B (en) | Image-scaling method and system based on edge self-adaption | |
| JPH08205194A (en) | Device for prediction between motion compensated frames | |
| JP4569680B2 (en) | Resolution conversion method | |
| EP1092206A1 (en) | Method of accurately locating the fractional position of a template match point | |
| US9672590B2 (en) | Image interpolation method and device based on autoregressive model | |
| US20170330311A1 (en) | Image processing device and method, image capturing device, program, and record medium | |
| CN112508783B (en) | Image processing method based on direction interpolation, computer device and computer readable storage medium | |
| CN107123086B (en) | Image scaling method, image scaling device and electronic equipment | |
| US8295646B2 (en) | Resolution converting method | |
| CN114049268A (en) | Image correction method, image correction device, electronic equipment and computer-readable storage medium | |
| CN109978898B (en) | Contour detection method based on vector field energy calculation | |
| US20120038785A1 (en) | Method for producing high resolution image | |
| US10397514B2 (en) | Projection apparatus, image processing apparatus, and image processing method | |
| KR101556625B1 (en) | Method for interpolating image magnification | |
| JP4487179B2 (en) | Image processing apparatus and method, recording medium, and program | |
| CN106611405B (en) | Image interpolation method and device | |
| KR20190023846A (en) | Method and apparatus for content-based image resizing | |
| JP2007035025A (en) | Image conversion apparatus, method thereof, program and computer-readable storage medium | |
| CN111093045B (en) | Method and device for scaling video sequence resolution | |
| US20100080488A1 (en) | Fast directional image interpolator with difference projection | |
| JP2017201454A (en) | Image processing device and program | |
| CN112836821B (en) | Convolution neural network calculated quantity compression method for automatic driving |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: A data augmentation method for the design of self-supporting steel chimneys Effective date of registration: 20230625 Granted publication date: 20230512 Pledgee: Weihai commercial bank Limited by Share Ltd. Qingdao branch Pledgor: QINGDAO HAOYU HEAVY INDUSTRY Co.,Ltd. Registration number: Y2023980045562 |