CN116596810B - Automatic enhancement method for spine endoscope image - Google Patents
Automatic enhancement method for spine endoscope image Download PDFInfo
- Publication number
- CN116596810B CN116596810B CN202310868769.4A CN202310868769A CN116596810B CN 116596810 B CN116596810 B CN 116596810B CN 202310868769 A CN202310868769 A CN 202310868769A CN 116596810 B CN116596810 B CN 116596810B
- Authority
- CN
- China
- Prior art keywords
- connected domain
- radius
- initial
- pixel points
- edge pixel
- 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
- 238000000034 method Methods 0.000 title claims abstract description 52
- 238000003708 edge detection Methods 0.000 claims abstract description 35
- 238000004891 communication Methods 0.000 claims description 11
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000012216 screening Methods 0.000 claims description 4
- 230000000694 effects Effects 0.000 abstract description 8
- 238000012545 processing Methods 0.000 abstract description 4
- 230000003313 weakening effect Effects 0.000 abstract description 2
- 230000008569 process Effects 0.000 description 11
- 238000010586 diagram Methods 0.000 description 5
- 238000003384 imaging method Methods 0.000 description 4
- 238000006243 chemical reaction Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000002708 enhancing effect Effects 0.000 description 3
- 238000010521 absorption reaction Methods 0.000 description 2
- 230000015572 biosynthetic process Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 239000003086 colorant Substances 0.000 description 2
- 201000010099 disease Diseases 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/40—Image enhancement or restoration using histogram techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/187—Segmentation; Edge detection involving region growing; involving region merging; involving connected component labelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10068—Endoscopic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10116—X-ray image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30008—Bone
- G06T2207/30012—Spine; Backbone
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Geometry (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of image processing, in particular to an automatic enhancement method of a spine endoscope image, which comprises the following steps: obtaining an edge detection image according to the gray level image; obtaining a connected domain according to the edge detection image; obtaining an index of the number of edge pixel points according to the connected domain; obtaining the maximum theoretical radius according to the Frieman chain code; obtaining the minimum circle radius according to the Euclidean distance; obtaining a linear degree index according to the minimum bounding circle radius and the theoretical maximum radius; obtaining a target degree according to the number index and the linear degree index of the edge pixel points; obtaining an initial connected domain according to the target degree; obtaining correlation according to the initial connected domain; obtaining a spine initial connected domain pair according to the correlation; obtaining a foreground region and a background region according to the initial connected domain pair of the spine; and carrying out histogram equalization to different degrees according to the foreground region and the background region to obtain an enhanced image. The invention further realizes the enhancement of the foreground region and the weakening of the background region, thereby achieving better enhancement effect.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to an automatic enhancement method for an image of a spinal endoscope.
Background
The X-ray imaging technology is a clinically common imaging auxiliary examination means at present, wherein the X-rays used belong to electromagnetic waves. Because the X-ray can penetrate human tissue, and different tissues of the human body have different absorption degrees on the X-ray, the X-ray can show the characteristics of images with different colors, so the X-ray detector is used for detecting and diagnosing various clinical diseases. In addition, the X-ray image is often affected by various factors during the formation process, and the quality of the X-ray image is reduced, such as blurring and darkness, so that the image enhancement of the X-ray image is required.
The common image enhancement method is global histogram equalization, and because the contrast distribution of a target area and a non-target area in an X-ray image is unbalanced, the global histogram equalization of the whole image often excessively enhances the contrast of the non-target area, so that the details of the target area cannot achieve an ideal enhancement effect. For the spine endoscope image, the spine region is a target region, the invention carries out edge detection on the X-ray image by an image processing technology, and carries out histogram equalization enhancement to different degrees on a foreground region and a background region which are segmented out of the spine region according to the spine region characteristics, thereby achieving the purpose of enhancing the image precision.
Disclosure of Invention
The invention provides an automatic enhancement method for a spinal endoscope image, which aims to solve the existing problems.
The invention relates to an automatic enhancement method of a spinal endoscope image, which adopts the following technical scheme:
an embodiment of the invention provides an automatic spine endoscope image enhancement method, which comprises the following steps:
acquiring a backbone X-ray image, and graying the backbone X-ray image to obtain a gray image;
performing edge detection according to the gray level image to obtain an edge detection image; obtaining a connected domain according to the edge detection image; obtaining an index of the number of edge pixel points of each connected domain according to the number of the pixel points of the connected domain; carrying out direction statistics on the connected domains according to the Frieman chain code to obtain the maximum theoretical radius of each connected domain; obtaining the linear degree index of each connected domain according to the minimum enclosing circle radius and the theoretical maximum radius; obtaining the target degree of each connected domain according to the number index and the linear degree index of the edge pixel points;
threshold screening is carried out on each connected domain according to the target degree to obtain an initial connected domain; obtaining the correlation between every two initial connected domains according to the initial connected domains; threshold screening is carried out on the initial connected domain according to the correlation to obtain a spine initial connected domain pair; obtaining a foreground region and a background region in the gray level image according to the initial connected domain pair;
and carrying out histogram equalization to different degrees according to the foreground region and the background region to obtain an enhanced image.
Preferably, the method for obtaining the edge pixel point number index of each connected domain according to the pixel point number of the connected domain specifically includes:
all pixel points with gray values not being 0 in the edge detection image are marked as edge pixel points, and the number of the edge pixel points in each connected domain is marked as a first number; recording the calculation result of the first quantity as the independent variable of the arc tangent trigonometric function as a first result; combine the first result withThe ratio result of (2) is recorded as the number index of the edge pixel points of each connected domain.
Preferably, the direction statistics is performed on the connected domains according to the frieman chain code to obtain the maximum theoretical radius of each connected domain, and the specific acquisition method is as follows:
if the Frieman chain code has even number and odd number, all edge pixel points with even number are marked as edge pixel points distributed along the 0 direction, all edge pixel points with odd number are marked as edge pixel points distributed along the 3 direction, and half distance length of the maximum Euclidean distance formed by the edge pixel points distributed along the two directions is marked as the maximum theoretical radius of each connected domain; if the Frieman chain code only has even number, the edge pixel points with even number are marked as edge pixel points distributed along the 0 direction, and the half distance length of the maximum Euclidean distance formed by the edge pixel points is marked as the maximum theoretical radius of each connected domain; if the Frieman chain code only has an odd number value, the edge pixel points with the odd number value are marked as edge pixel points distributed along the 3 direction, and the half distance length of the maximum Euclidean distance formed by the edge pixel points is marked as the maximum theoretical radius of each connected domain, so that the maximum theoretical radius of each connected domain is obtained.
Preferably, the linear degree index of each connected domain is obtained according to the radius of the minimum enclosing circle and the theoretical maximum radius, and the specific obtaining method comprises the following steps:
and (3) recording the ratio of the radius of the minimum enclosing circle of each connected domain to the maximum theoretical radius as the linear degree index of each connected domain.
Preferably, the minimum circle radius of each connected domain is obtained by the following specific method:
the corresponding two edge pixel points when the Euclidean distance in each connected domain takes the maximum value are marked as an initial pixel point pair, and the half distance length of the corresponding Euclidean distance is marked as a first radius; then, the midpoint on the first diameter is marked as an initial center point; then, marking an edge pixel point with the largest Euclidean distance formed between each connected domain and the initial center point as a first outer point, and marking the corresponding Euclidean distance as a second radius; if the second radius is larger than the first radius, taking the initial circle center point as the minimum enclosing circle center, and recording a circle with the second radius being the minimum enclosing circle radius as the minimum enclosing circle of each communication domain; and if the second radius is smaller than or equal to the first radius, taking the initial center point as the minimum bounding center, and recording the circle with the first radius being the minimum bounding circle radius as the minimum bounding circle of each communication domain to obtain the minimum bounding circle radius of each communication domain.
Preferably, the target degree of each connected domain is obtained according to the number index and the linear degree index of the edge pixel points, and the specific obtaining method comprises the following steps:
multiplying the linear degree index of each connected domain by the edge pixel point quantity index, and recording the obtained multiplication result as the target degree of each connected domain.
Preferably, the obtaining the correlation between every two initial connected domains according to the initial connected domains specifically includes:
the multiplication result of the target degree of the jth initial connected domain and the jth+1th initial connected domain is recorded as a first product; the summation result of the number of the edge pixel points of the jth initial connected domain and the jth+1th initial connected domain is recorded as a second number; recording the result of the first product and the second number as a second product; the Euclidean distance between the j-th initial communicating domain and the j+1th initial communicating domain at the two nearest points is recorded as a first denominator; and (3) marking the ratio of the second product to the first denominator as the correlation between the j-th initial connected domain and the j+1-th initial connected domain, and obtaining the correlation between any two initial two connected domains in all the initial connected domains.
The technical scheme of the invention has the beneficial effects that: by analyzing the edge continuity and linearity in the edge detection image, the image is segmented with a smaller calculation amount to obtain a more accurate spine edge. The enhancement of the foreground region and the weakening of the background region in the X-ray image are realized by carrying out histogram equalization on the foreground region and the background region to different degrees, so that a better enhancement effect is achieved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of steps of an automatic enhancement method for an endoscopic image of the present invention;
FIG. 2 is a gray scale image of a spinal X-ray image of the present invention;
FIG. 3 is an edge-detected image of a spinal X-ray image of the present invention at an edge-detected intensity;
FIG. 4 is a schematic diagram of the linear degree of the present invention;
FIG. 5 is a schematic diagram of the Frieman chain code direction sequence of the present invention;
fig. 6 is a schematic diagram of the maximum theoretical radius of the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following description refers to the specific implementation, structure, characteristics and effects of an automatic enhancement method for a spinal endoscope image according to the invention by combining the accompanying drawings and the preferred embodiment. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
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 automatic enhancement method for the spine endoscope image provided by the invention with reference to the accompanying drawings.
Referring to fig. 1, a flowchart of steps of a method for automatically enhancing an image of a spinal endoscope according to an embodiment of the invention is shown, the method includes the following steps:
step S001: and acquiring a backbone X-ray image, and graying the backbone X-ray image to obtain a gray image.
It should be noted that the X-ray imaging technique is a clinically common imaging auxiliary inspection method at present, wherein the used X-ray belongs to an electromagnetic wave. Because the X-ray can penetrate human tissue, and different tissues of the human body have different absorption degrees on the X-ray, the X-ray can show the characteristics of images with different colors, so the X-ray detector is used for detecting and diagnosing various clinical diseases. In addition, the X-ray image is often affected by various factors during the formation process, and the quality of the X-ray image is reduced, such as blurring and darkness, so that the image enhancement of the X-ray image is required.
The common image enhancement method is global histogram equalization, and because the contrast distribution of a target area and a non-target area in an X-ray image is unbalanced, the global histogram equalization of the whole image often excessively enhances the contrast of the non-target area, so that the details of the target area cannot achieve an ideal enhancement effect. For spinal endoscopic images, the spinal region is often the portion of greater concern, the target region. According to the embodiment, the edge detection is carried out on the X-ray image through the image processing technology, the spine region is segmented according to the spine region characteristics to obtain the foreground region and the background region, and then the histogram equalization enhancement is respectively carried out on the foreground region and the background region to different degrees, so that the purpose of accurately enhancing the image is achieved. The X-ray image can be a digitized image with bitmap conversion generally in the form of an interface screenshot. The obtained digitized image is subjected to graying to obtain a gray image, and the gray image is converted into the gray image, so that the memory consumption can be greatly reduced, the operation is simplified, and the subsequent image operation is carried out on the gray image.
Specifically, the specific process of acquiring the gray level image of the spine X-ray image is as follows: an experimenter shoots by using an X-ray machine, captures corresponding X-ray image data, then transmits the data to a computer system to generate an original image of a spine X-ray image, and then obtains a gray level image of the spine X-ray image through linear gray level conversion, and referring to FIG. 2, the gray level image of the spine X-ray image is shown.
Thus, a gray scale image of the spine X-ray image is obtained.
Step S002: and carrying out edge detection on the gray level image, and obtaining the target degree according to the connectivity of the image edge.
It should be noted that, since the spine region of the spine endoscope image has a more obvious edge region range in the gray level image, the edge detection is performed on the gray level image, so that a relatively obvious spine region can be obtained, and other regions do not have obvious edges like the spine region. And because the spine region is relatively complete relative to the edge region parts of other regions, the connectivity between the edge pixel points of the spine region edge region is better, and therefore, a relatively complete spine region segmentation image can be obtained through the connectivity between the edge pixel points in the edge detection image and the correlation between the connected regions.
1. And obtaining all connected domains of the edge detection image according to the gray level image.
The core size is smaller than that of other coresThe edge detection operator, roberts operator, uses +.>The kernel of the X-ray image is used for edge detection, is more sensitive to edge change, is more suitable for an image with more gentle gray level change, such as an X-ray image, and uses +.>The operator core of (2) can greatly reduce the calculated amount, obtain better algorithm effect, and the Roberts operator is sensitive to detail reaction, and the calculation is usedEdge detection can better preserve various edges in the image.
Specifically, the specific process of acquiring the edge detection image and all the corresponding connected domains is as follows: an edge detection intensity T1 is preset, where the embodiment is described by taking t1=0.05 as an example, and the embodiment is not limited specifically, where T1 may be determined according to the specific implementation. The edge detection is performed on the gray image by the Roberts operator with the edge detection intensity T1 to obtain an edge detection image corresponding to the gray image, please refer to fig. 3, which illustrates an edge detection image of the spine X-ray image under an edge detection intensity, that is, the edge detection intensity is set to 0.05 in an embodiment of the present invention. And then carrying out flood filling on the edge detection image to obtain all connected domains of the edge detection image.
It should be noted that, the Roberts operator edge detection and the flood filling method are both the prior art, and this embodiment will not be described.
So far, all connected domains of the edge detection image can be obtained through the method.
2. And obtaining the connected domain belonging to the spine region according to the number index of the edge pixel points in the connected domain and the linearity degree of the connected domain.
In all the connected regions of the acquired edge detection image, the number of edge pixels in the connected regions belonging to the spine region is large and the shape of the connected regions is close to a linear shape, whereas the number of edge pixels in the connected regions not belonging to the spine region is small and the shape of the connected regions is expressed in an irregular shape. Therefore, the connected domain belonging to the spine region can be obtained according to the number of the edge pixel points in the connected domain and the shape of the connected domain.
1) And obtaining an index of the number of the edge pixel points.
Specifically, the specific process for obtaining the number index of the edge pixel points in the connected domain is as follows: all pixel points with gray values not being 0 in the edge detection image are marked as edge pixel points, and the number of the edge pixel points in each connected domain is marked as a first number; the result of the calculation of the first quantity as the argument of the arctangent trigonometric function is recorded as a firstResults; combine the first result withThe ratio result of (2) is recorded as the number index of the edge pixel points of each connected domain.
Thus, the number index of the edge pixel points of each connected domain can be obtained through the method.
2) Obtaining the linear degree index.
It should be noted that, only when the number index of the edge pixel points of the connected domain reaches a certain degree, the shape of the corresponding connected domain is more likely to be a spine region as the shape of the corresponding connected domain is closer to a line shape; conversely, noise region interference is most likely. Referring to fig. 4, a schematic diagram of the degree of linearity is shown.
In fig. 4, when the number of edge pixels included in two connected regions is the same, the radius of the minimum bounding circle formed by the edge pixels increases as the edge pixel distribution approaches a straight line; conversely, the smaller the radius. The degree of linearity of each connected domain can be obtained according to the radius of the minimum bounding circle.
It should be further noted that, for each connected domain in the image, the theoretical maximum radius is obtained by analyzing the connection condition and calculating. The Frieman chain code is an edge chain code which is small in required calculation amount and can well reflect the edge communication condition, and the communication condition inside the edge is obtained by extracting the Frieman chain code of the edge. Referring to fig. 5, a schematic diagram of the frieman chain code direction sequence is shown; referring to fig. 6, a maximum theoretical radius schematic is shown.
In fig. 5, the whole connected domain is uniformly divided into eight directions, namely, numerals 0 to 7 in fig. 5 respectively represent the corresponding eight directions, wherein an odd number value represents edge pixel points distributed obliquely, and an even number value represents edge pixel points distributed longitudinally or transversely. In fig. 6, co represents the number of odd numbers in the frieman chain code, ce represents the number of even numbers in the frieman chain code, and the maximum theoretical radius can be obtained by constructing a triangle by the number of odd numbers and even numbers in the frieman chain code as shown in fig. 6.
Specifically, describing the ith connected domain as the current connected domain, and acquiring the specific process of the maximum theoretical radius of the connected domain: if the Frieman chain code has even number and odd number, all edge pixel points with even number are marked as edge pixel points distributed along the 0 direction, all edge pixel points with odd number are marked as edge pixel points distributed along the 3 direction, and half distance length of the maximum Euclidean distance formed by the edge pixel points distributed along the two directions is marked as the maximum theoretical radius of the current connected domain; if the Frieman chain code only has even number, marking the edge pixel points with the even number as edge pixel points distributed along the 0 direction, and marking the half distance length of the maximum Euclidean distance formed by the edge pixel points as the maximum theoretical radius of the current connected domain; if the Frieman chain code only has an odd number value, the edge pixel points with the odd number value are marked as edge pixel points distributed along the 3 direction, and the half distance length of the maximum Euclidean distance formed by the edge pixel points is marked as the maximum theoretical radius of the current connected domain. In addition, the euclidean distance and the frieman chain code are related art, and this embodiment will not be described.
So far, the maximum theoretical radius of the current connected domain is obtained through the method, and the maximum theoretical radius of each connected domain is obtained.
Further, the specific process of obtaining the minimum surrounding circle of the connected domain is as follows: marking two corresponding edge pixel points as initial pixel point pairs when the Euclidean distance in the current connected domain takes the maximum value, and marking half of the distance length of the corresponding Euclidean distance as a first radius; then, the midpoint on the first diameter is marked as an initial center point; then, marking an edge pixel point with the largest Euclidean distance formed between the current connected domain and the initial center point as a first outer point, and marking the corresponding Euclidean distance as a second radius; if the second radius is larger than the first radius, taking the initial circle center point as the minimum enclosing circle center, and marking the circle with the second radius being the minimum enclosing circle radius as the minimum enclosing circle of the current communication domain; and if the second radius is smaller than or equal to the first radius, taking the initial circle center point as the minimum bounding circle center, and recording the circle with the first radius being the minimum bounding circle radius as the minimum bounding circle of the current communication domain.
So far, the minimum bounding circle and the minimum bounding circle radius of the current communicating domain are obtained through the method, and the minimum bounding circle radius of each communicating domain is obtained.
Specifically, the specific process for obtaining the linear degree index of the current connected domain according to the minimum enclosing circle radius and the maximum theoretical radius of the current connected domain is as follows: and (3) recording the ratio of the radius of the minimum enclosing circle of the current connected domain to the maximum theoretical radius as the linear degree index of the current connected domain.
Thus, the linear degree index of the current connected domain is obtained through the method, and the linear degree index of each connected domain is obtained.
Further, if the index of the degree of alignment of the ith communicating region is large, it is only possible to explain that the shape of the current communicating region is approaching alignment. On the basis, if the index of the number of the edge pixel points of the current connected domain is smaller, the number of the edge pixel points contained in the current connected domain is smaller, and the current connected domain is possibly a connected domain of the spine region but is more likely to be a noise point region; if the index of the number of the edge pixel points of the current connected domain is larger, the number of the edge pixel points contained in the current connected domain is larger, the possibility that the current connected domain is a connected domain of the spine region is larger, and the possibility that the current connected domain belongs to a connected domain of other regions is smaller. Therefore, the target degree of the connected domain belonging to the spine region needs to be obtained according to the number index of the edge pixel points of the connected domain.
Specifically, the specific process for obtaining the target degree of the connected domain belonging to the spine region is as follows: multiplying the linear degree index of the current connected domain by the edge pixel point quantity index, and recording the obtained multiplication result as the target degree of the current connected domain.
Thus, the target degree of the current connected domain is obtained through the method, and the target degree of each connected domain is obtained.
Step S003: and obtaining an initial connected domain according to the target degree, and obtaining a spine region according to the initial connected domain.
It should be noted that, according to the target degree in step S002, the initial connected domain belonging to the spine region may be obtained preliminarily. In the initial connected domains, most of the initial connected domains are distributed more tightly, and a small part of the initial connected domains are distributed more dispersedly, wherein the initial connected domains distributed more tightly are target connected domains, and the initial connected domains distributed more dispersedly are interference connected domains, so that the spine connected domain truly belonging to the spine region can be obtained by analyzing the correlation of the initial connected domains according to the distance between the initial connected domains.
Specifically, a target level threshold T2 is preset, where the embodiment is described by taking t2=0.5 as an example, and the embodiment is not specifically limited, where T2 may be determined according to the specific implementation situation. If the target degree of the current connected domain is greater than T2, the current connected domain is the initial connected domain.
So far, all initial connected domains in the edge detection image are obtained through the method.
Further, describing the j-th initial connected domain and the j+1th initial connected domain as examples, the specific process of obtaining the correlation between the two initial connected domains is as follows: the multiplication result of the target degree of the jth initial connected domain and the jth+1th initial connected domain is recorded as a first product; the summation result of the number of the edge pixel points of the jth initial connected domain and the jth+1th initial connected domain is recorded as a second number; recording the result of the first product and the second number as a second product; the Euclidean distance between the two nearest edge pixel points between the jth initial connected domain and the (j+1) th initial connected domain is recorded as a first denominator; and (3) marking the ratio of the second product to the first denominator as the correlation between the j-th initial connected domain and the j+1-th initial connected domain, and obtaining the correlation between any two initial connected domains in all the initial connected domains. A correlation threshold T3 is preset, where the present embodiment is described by taking t3=1 as an example, and the present embodiment is not limited specifically, where T3 may be determined according to the specific implementation situation. If the correlation between the jth initial communicating region and the jth+1th initial communicating region is greater than T3, the jth initial communicating region and the jth+1th initial communicating region are a pair of spine initial communicating regions.
So far, all the initial connected domain pairs of the spine in the edge detection image are obtained through the method.
The union set formed by all the pairs of the initial connected regions of the spine is referred to as a first union set, and all the initial connected regions included in the first union set are referred to as spine connected regions.
Thus, the spine region surrounded by the spine communication region is obtained by the method.
Step S004: and obtaining an image foreground region and an image background region according to the spine region, and carrying out histogram equalization to different degrees to obtain an enhanced image.
It should be noted that, in the step S003, the spine region has a corresponding region in the gray level image, and the region is a region to be enhanced, so that in order to improve the gray level comparison of the spine region, the display effect of the background region is compressed, and the spine region and the background region can be enhanced to different degrees by adjusting the gray level range of histogram equalization.
Specifically, the specific process of adjusting histogram equalization is as follows: the corresponding region of the obtained spine region in the gray level image is marked as a foreground region, and the rest region in the gray level image is marked as a background region; adjusting the gray value range of the histogram equalization of the foreground region to be 1.5 times of the initial gray value range, and then carrying out the local histogram equalization after the gray value range adjustment in the foreground region; and adjusting the gray value range of the histogram equalization of the background area to be 0.8 times of the initial gray value range, and then carrying out the local histogram equalization after adjusting the gray value range in the background area.
So far, the enhanced image is obtained by the above method.
This embodiment is completed.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, alternatives, and improvements that fall within the spirit and scope of the invention.
Claims (7)
1. An automatic enhancement method for a spinal endoscope image is characterized by comprising the following steps:
acquiring a backbone X-ray image, and graying the backbone X-ray image to obtain a gray image;
performing edge detection according to the gray level image to obtain an edge detection image; obtaining a connected domain according to the edge detection image; obtaining an index of the number of edge pixel points of each connected domain according to the number of the pixel points of the connected domain; carrying out direction statistics on the connected domains according to the Frieman chain code to obtain the maximum theoretical radius of each connected domain; obtaining the linear degree index of each connected domain according to the minimum enclosing circle radius and the theoretical maximum radius; obtaining the target degree of each connected domain according to the number index and the linear degree index of the edge pixel points;
threshold screening is carried out on each connected domain according to the target degree to obtain an initial connected domain; obtaining the correlation between every two initial connected domains according to the initial connected domains; threshold screening is carried out on the initial connected domain according to the correlation to obtain a spine initial connected domain pair; obtaining a foreground region and a background region in the gray level image according to the initial connected domain pair;
and carrying out histogram equalization to different degrees according to the foreground region and the background region to obtain an enhanced image.
2. The automatic enhancement method of the spine endoscope image according to claim 1, wherein the obtaining the index of the number of the edge pixels of each connected domain according to the number of the pixels of the connected domain comprises the following steps:
all pixel points with gray values not being 0 in the edge detection image are marked as edge pixel points, and the number of the edge pixel points in each connected domain is marked as a first number; recording the calculation result of the first quantity as the independent variable of the arc tangent trigonometric function as a first result; combine the first result withThe ratio result of (2) is recorded as the number index of the edge pixel points of each connected domain.
3. The automatic enhancement method of the spine endoscope image according to claim 1, wherein the direction statistics is performed on the connected domains according to the frieman chain code to obtain the maximum theoretical radius of each connected domain, and the specific acquisition method is as follows:
if the Frieman chain code has even number and odd number, all edge pixel points with even number are marked as edge pixel points distributed along the 0 direction, all edge pixel points with odd number are marked as edge pixel points distributed along the 3 direction, and half distance length of the maximum Euclidean distance formed by the edge pixel points distributed along the two directions is marked as the maximum theoretical radius of each connected domain; if the Frieman chain code only has even number, the edge pixel points with even number are marked as edge pixel points distributed along the 0 direction, and the half distance length of the maximum Euclidean distance formed by the edge pixel points is marked as the maximum theoretical radius of each connected domain; if the Frieman chain code only has an odd number value, the edge pixel points with the odd number value are marked as edge pixel points distributed along the 3 direction, and the half distance length of the maximum Euclidean distance formed by the edge pixel points is marked as the maximum theoretical radius of each connected domain, so that the maximum theoretical radius of each connected domain is obtained.
4. The automatic enhancement method of the spine endoscope image according to claim 1, wherein the linear degree index of each connected domain is obtained according to the radius of the minimum enclosing circle and the theoretical maximum radius, and the specific acquisition method comprises the following steps:
and (3) recording the ratio of the radius of the minimum enclosing circle of each connected domain to the maximum theoretical radius as the linear degree index of each connected domain.
5. The automatic enhancement method of the spine endoscope image according to claim 4, wherein the minimum circle radius of each connected domain is obtained by the following steps:
the corresponding two edge pixel points when the Euclidean distance in each connected domain takes the maximum value are marked as an initial pixel point pair, and the half distance length of the corresponding Euclidean distance is marked as a first radius; then, the midpoint on the first diameter is marked as an initial center point; then, marking an edge pixel point with the largest Euclidean distance formed between each connected domain and the initial center point as a first outer point, and marking the corresponding Euclidean distance as a second radius; if the second radius is larger than the first radius, taking the initial circle center point as the minimum enclosing circle center, and recording a circle with the second radius being the minimum enclosing circle radius as the minimum enclosing circle of each communication domain; and if the second radius is smaller than or equal to the first radius, taking the initial center point as the minimum bounding center, and recording the circle with the first radius being the minimum bounding circle radius as the minimum bounding circle of each communication domain to obtain the minimum bounding circle radius of each communication domain.
6. The automatic enhancement method of the spine endoscope image according to claim 1, wherein the target degree of each connected domain is obtained according to the number index and the linear degree index of the edge pixel points, and the specific acquisition method comprises the following steps:
multiplying the linear degree index of each connected domain by the edge pixel point quantity index, and recording the obtained multiplication result as the target degree of each connected domain.
7. The automatic enhancement method of a spinal endoscope image according to claim 1, wherein the obtaining the correlation between every two initial connected domains according to the initial connected domains comprises the following specific steps:
the multiplication result of the target degree of the jth initial connected domain and the jth+1th initial connected domain is recorded as a first product; the summation result of the number of the edge pixel points of the jth initial connected domain and the jth+1th initial connected domain is recorded as a second number; recording the result of the first product and the second number as a second product; the Euclidean distance between the j-th initial communicating domain and the j+1th initial communicating domain at the two nearest points is recorded as a first denominator; and (3) marking the ratio of the second product to the first denominator as the correlation between the j-th initial connected domain and the j+1-th initial connected domain, and obtaining the correlation between any two initial two connected domains in all the initial connected domains.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310868769.4A CN116596810B (en) | 2023-07-17 | 2023-07-17 | Automatic enhancement method for spine endoscope image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310868769.4A CN116596810B (en) | 2023-07-17 | 2023-07-17 | Automatic enhancement method for spine endoscope image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116596810A CN116596810A (en) | 2023-08-15 |
| CN116596810B true CN116596810B (en) | 2023-09-19 |
Family
ID=87601247
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310868769.4A Active CN116596810B (en) | 2023-07-17 | 2023-07-17 | Automatic enhancement method for spine endoscope image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116596810B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117893530B (en) * | 2024-03-14 | 2024-06-04 | 中国医科大学附属第一医院 | Throat image analysis system based on artificial intelligence |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110770636A (en) * | 2017-04-25 | 2020-02-07 | 雷特克斯有限公司 | Wearable image processing and control system with functions of correcting visual defects, enhancing vision and sensing ability |
| WO2021031459A1 (en) * | 2019-08-16 | 2021-02-25 | 域鑫科技（惠州）有限公司 | Image enhancement method and apparatus suitable for endoscope, and storage medium |
-
2023
- 2023-07-17 CN CN202310868769.4A patent/CN116596810B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110770636A (en) * | 2017-04-25 | 2020-02-07 | 雷特克斯有限公司 | Wearable image processing and control system with functions of correcting visual defects, enhancing vision and sensing ability |
| WO2021031459A1 (en) * | 2019-08-16 | 2021-02-25 | 域鑫科技（惠州）有限公司 | Image enhancement method and apparatus suitable for endoscope, and storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 基于目标特征的灰度直方图重分布图象增强算法;吴燕;张晨燕;;计算机与信息技术(08);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116596810A (en) | 2023-08-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US7646902B2 (en) | Computerized detection of breast cancer on digital tomosynthesis mammograms | |
| US9002134B2 (en) | Multi-scale image normalization and enhancement | |
| EP1316919A2 (en) | Method for contrast-enhancement of digital portal images | |
| JPH09508814A (en) | Automatic method and system for segmenting medical images | |
| WO2006024974A1 (en) | Feature weighted medical object contouring using distance coordinates | |
| CN115496771A (en) | Brain tumor segmentation method based on brain three-dimensional MRI image design | |
| CN116596810B (en) | Automatic enhancement method for spine endoscope image | |
| KR20230059799A (en) | A Connected Machine Learning Model Using Collaborative Training for Lesion Detection | |
| US20100220913A1 (en) | System and Method for Detection of Lesions in Three-Dimensional Digital Medical Image | |
| CN117422628B (en) | Optimized enhancement method for cardiac vascular ultrasonic examination data | |
| CN112690813A (en) | Fractal theory-based tumor heterogeneity analysis method, system, terminal and medium | |
| CN112614093A (en) | Breast pathology image classification method based on multi-scale space attention network | |
| George et al. | Preprocessing filters for mammogram images: a review | |
| CN116309647B (en) | Method for constructing craniocerebral lesion image segmentation model, image segmentation method and device | |
| Kipele et al. | Poisson noise reduction with nonlocal-pca hybrid model in medical x-ray images | |
| CN111062909A (en) | Method and equipment for judging benign and malignant breast tumor | |
| CN112634280B (en) | MRI image brain tumor segmentation method based on energy functional | |
| CN113689424B (en) | Ultrasonic inspection system capable of automatically identifying image features and identification method | |
| Bocchi et al. | Segmentation of ultrasound breast images: optimization of algorithm parameters | |
| CN113920100A (en) | Knowledge distillation-based weak supervision bone scanning image hot spot segmentation method and system | |
| Noviana et al. | Axial segmentation of lungs CT scan images using canny method and morphological operation | |
| CN113768544A (en) | Ultrasonic imaging method and equipment for mammary gland | |
| Salehi et al. | Investigation and simulation of different medical image processing algorithms to improve image quality using simulink matlab | |
| Gui et al. | Liver segmentation for CT images using an improved GGVF-snake | |
| CN117274130B (en) | Self-adaptive enhancement method for oral cancer medical image |
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 |