CN107610081B - A kind of endoscope signal processing method and system - Google Patents

Abstract

The invention relates to an endoscope signal processing method and system, which are characterized in that whether foreign matters exist on a lens is judged firstly, if foreign matters exist, a picture corresponding to the foreign matters of a main image in a picture group is captured and spliced with the main image, the spliced goodness of fit is detected, and a synthesized main image with the highest goodness of fit is selected as an output image. The method can reduce the influence of foreign matters on the lens on the endoscope picture, so that medical staff can obtain a more clear and comprehensive picture, and the diagnosis and treatment are facilitated.

Description

A kind of endoscope signal processing method and system

technical field

The invention relates to the technical field of endoscopes, in particular to an endoscope signal processing method and system.

Background technique

In recent years, surgical procedures using endoscopes such as laparoscopic surgery and thoracoscopic surgery have been attracting attention. Endoscopic surgery is advantageous because it does not require laparotomy, thoracotomy, etc., and only two or three holes of several centimeters in diameter need to be made for the insertion of endoscopes and surgical tools, thus significantly reducing the application of burden on patients. However, it is highly difficult to perform surgery with the extremely limited field of view of the endoscope, and foreign bodies in the patient's body will inevitably stick to the lens of the endoscope, and sometimes, the irrigation system of the endoscope will not be flushed out. The presence of a foreign body will cause the originally narrow vision to become more blurred, which will affect the doctor's judgment.

SUMMARY OF THE INVENTION

In order to overcome the above disadvantages, the present invention provides an endoscope signal processing method.

The technical scheme adopted in the present invention is: comprising the following steps: 1. Image generation, multiple images generated by photographing the time series of the subject; 2. The main and auxiliary images are determined, and an image is selected according to the static image generation instruction As the main image, multiple images are used as sub-image groups; 3. Contour extraction, extracting contour components from the main image and sub-images; 4. Judging foreign objects on the lens, determining whether there is foreign matter on the lens according to the relative position of the contour components in the image , if there is a foreign object, go to step 5, if there is no foreign object, determine the main image as the final image and go to step 9; 5. Select the features around the foreign object in the main image, select and record the features around the foreign object in the main image; 6. The corresponding part of the secondary image group Image extraction, determine the corresponding position on the secondary image group of the part of the main image that is blocked by foreign objects according to the surrounding features of the foreign object in the main image, and extract the image of the corresponding position; 7. The image of the corresponding position is synthesized with the main image, and the corresponding position of the image is synthesized. The images are spliced to the foreign objects of the main image, and multiple composite main images are generated; 8. The texture of the composite main image is judged, the degree of agreement of the main image synthesis place is detected, and the one with the highest degree of agreement among the multiple composite main images is selected as the final image. 9. Save and output the final image, save multiple composite main images, main images and sub-image groups.

The present invention also provides an endoscope signal processing system, comprising:

An image generation module, which generates multiple images through time-series imaging of the subject;

The main and auxiliary image determination module selects one image as the main image and multiple images as the auxiliary image group according to the static image generation instruction;

a contour extraction module, extracting contour components from the main image and the sub-image;

The foreign object judgment module determines whether there is a foreign object on the lens according to the relative position of the contour components in the image. If there is a foreign object, activate the feature selection module around the foreign object in the main image. If there is no foreign object, determine the main image as the final image and activate the save output module;

a feature selection module around the foreign body in the main image, which selects and records the features around the foreign body in the main image;

an image extraction module corresponding to the sub-image group, determining the corresponding position of the part of the main image that is occluded by the foreign object on the sub-image group according to the surrounding features of the foreign object in the main image, and extracting the image of the corresponding position;

A synthetic main image module, which splices the images at the corresponding positions to the foreign objects of the main image to generate multiple synthetic main images;

a texture discrimination module, which detects the degree of conformity at the synthesis place of the main image, and selects the one with the highest degree of conformity among the plurality of synthesized main images as the final image;

The saving and outputting module saves and outputs the final image, and saves the plurality of composite main images, the main image and the sub-image group.

The effects of the present invention are: the endoscope signal processing method and system described in the present invention can reduce the influence of foreign objects on the lens on the endoscope picture, so that medical staff can obtain a clearer and more comprehensive picture, which is helpful for The conduct of diagnosis and treatment.

Description of drawings

FIG. 1 is a flowchart of an endoscope signal processing method provided by the present invention;

FIG. 2 shows a structural block diagram of an endoscope signal processing system provided by the present invention;

Fig. 3 shows the structural block diagram of the main and auxiliary image determination module in Fig. 2;

Fig. 4 shows the structural block diagram of the foreign object judgment module in Fig. 2;

Fig. 5 shows the structural block diagram of the image extraction module corresponding to the sub-image group in Fig. 2;

Fig. 6 shows the structural block diagram of the texture discrimination module in Fig. 2;

In the figure: 1-endoscope signal processing system, 11-image generation module, 12-main and auxiliary image determination module, 121-command receiving module, 122-main image selection module, 123-sub-image selection module, 13- Contour extraction module, 14—foreign object judgment module, 141—contour shape comparison module, 142—contour position comparison module, 15—main image feature selection module around foreign objects, 16—image extraction module corresponding to the pair of image groups, 161—comparison Feature module, 162-image interception module, 17-synthesis main image module, 18-texture discrimination module, 181-contour re-extraction module, 182-expansion processing module, 183-fitness detection module, 184-fitness comparison module, 19 - Save the output module.

Detailed ways

The following describes an endoscope signal processing method provided by the present invention in conjunction with the accompanying drawings:

Please refer to FIG. 1, which is an endoscope signal processing method provided by the present invention, including the steps in the following order:

S1, image generation, a plurality of images generated by time-series imaging of the subject;

S2, the main and auxiliary images are determined, and one image is selected as the main image according to the static image generation instruction, and multiple images are used as the auxiliary image group;

S3, contour extraction, extracting contour components from the main image and the sub-image;

S4. Judgment of foreign objects on the lens, according to the relative position of the contour components in the image to determine whether there is a foreign object on the lens, if there is a foreign object, execute step S5, if there is no foreign matter, determine that the main image is the final image and execute step S9;

S5. Select the features around the foreign body in the main image, select and record the features around the foreign body in the main image;

S6, extracting images corresponding to the sub-image group, determining the corresponding position of the part of the main image that is blocked by the foreign object on the sub-image group according to the surrounding features of the foreign object in the main image, and extracting the image of the corresponding position;

S7, the image of the corresponding position is synthesized with the main image, and the image of the corresponding position is spliced to the foreign body of the main image to generate multiple composite main images;

S8, synthesizing the main image texture discrimination, detecting the degree of agreement where the main image is synthesized, and selecting the one with the highest degree of agreement among the multiple synthesized main images as the final image;

S9, save and output the final image, and save a plurality of composite main images, main images and sub-image groups.

In the endoscope signal processing method according to the present invention, the step of determining the primary and secondary images in step S2 includes:

S21. Receive a static image generation instruction;

S22, receive the last frame of image before the static image generation instruction as the main image;

S23. Several frames of images before the last frame of image are used as a sub-picture group.

In the endoscope signal processing method according to the present invention, the step of judging foreign objects on the lens in step S4 includes:

S41, compare the shape of each contour component of the main image with each contour component on the sub-image group, if the contour component in the main image has the same shape as the contour component on the sub-image group, it is determined to be the same contour component;

S42. Measure and compare the lengths of the same contour component on the main image and the sub-image group from the edge of the main image and the sub-image group. If they are equal, it is determined that the same contour component is a foreign object.

In the endoscope signal processing method according to the present invention, the step of extracting images corresponding to the sub-image group in step S6 includes:

S61, searching for an image area with features around foreign objects in the main image on the secondary image group;

S62, intercept the image area.

In the endoscope signal processing method according to the present invention, the step of synthesizing the main image texture discrimination in step S8 includes:

S81. Re-extract the contour components at the synthetic splicing position of the synthetic main image;

S82, performing expansion processing on the contour components at the splicing position of the composite main image to generate an expanded composite main image;

S83, detecting the degree of fit of the lines at the splicing point of the dilated synthetic main image;

S84 , comparing the degrees of agreement of the plurality of dilated synthetic main images, and selecting a synthetic main image with the highest agreement as the final image.

Please refer to FIG. 2, the present invention also provides an endoscope signal processing system 1, including:

The image generation module 11 generates multiple images through time-series imaging of the subject;

The primary and secondary image determination module 12 selects one image as the primary image and multiple images as the secondary image group according to the static image generation instruction;

an outline extraction module 13, extracting outline components from the main image and the sub-image;

The foreign object determination module 14 determines whether there is a foreign object on the lens according to the relative position of the contour component in the image, if there is a foreign object, activates the feature selection module around the foreign object in the main image, if there is no foreign object, determines that the main image is the final image and activates the save output module;

The feature selection module 15 around the foreign object in the main image selects and records the features around the foreign object in the main image;

The image extraction module 16 corresponding to the sub-image group determines the corresponding position of the part of the main image that is blocked by the foreign object on the sub-image group according to the surrounding features of the foreign object in the main image, and extracts the image of the corresponding position;

The synthetic main image module 17 splices the images at the corresponding positions to the foreign objects in the main image to generate multiple synthetic main images;

The texture discrimination module 18 detects the degree of conformity at the synthesis place of the main image, and selects the one with the highest degree of conformity among the plurality of synthesized main images as the final image;

The saving and outputting module 19 saves and outputs the final image, and saves the plurality of composite main images, the main image and the sub-image group.

Please refer to FIG. 3, an endoscope signal processing system 1 according to the present invention, the main and auxiliary image determination module 12 includes:

The instruction receiving module 121 receives a static image generation instruction;

The main image selection module 122 selects the last frame image before receiving the static image generation instruction as the main image;

The sub-image selection module 123 selects several frames of images before the last frame of image as the sub-image group.

Please refer to FIG. 4 , an endoscope signal processing system 1 according to the present invention, the foreign object determination module 14 includes:

The contour shape comparison module 141 compares the shape of each contour component of the main image and each contour component on the sub-image group, and if the contour component of the main image has the same shape as the contour component on the sub-image group, it is determined to be the same contour component;

The contour position comparison module 142 measures and compares the length of the same contour component on the main image and the sub-image group from the edge of the main image and the sub-image group. If the lengths of the edges of the image groups are equal, it is determined that the same contour components are foreign objects.

Referring to FIG. 5, an endoscope signal processing system 1 according to the present invention, the image extraction module 16 corresponding to the sub-image group includes:

The feature comparison module 161 searches the sub-image group for an image area with features around foreign objects in the main image;

The image capturing module 162 captures the image area.

Please refer to FIG. 6 , an endoscope signal processing system 1 according to the present invention, the texture discrimination module 18 includes:

The contour re-extraction module 181, re-extracts the contour components at the synthetic splicing of the synthetic main image;

The expansion processing module 182 performs expansion processing on the contour components where the composite main image is spliced to generate an expanded composite main image;

The degree of fit detection module 183 detects the degree of fit of the lines at the splicing point of the expanded synthetic main image;

The matching degree comparison module 184 compares the matching degrees of the plurality of dilated composite main images, and selects a composite main image with the highest matching degree as the final image.

In the endoscope signal processing method and system of the present invention, it is first determined whether there is a foreign object on the lens, and if there is a foreign object, a picture corresponding to the foreign object in the main image on the image group is intercepted and spliced with the main image, and the splicing is detected. After the matching degree is determined, the composite main image with the highest matching degree is selected as the output image. The endoscope signal processing method and system can reduce the influence of foreign objects on the lens on the endoscope picture, so that the medical staff can obtain a clearer and more comprehensive picture, which is helpful for the diagnosis and treatment.

The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection of the present invention. within the range.

Claims (6)

1. An endoscope signal processing method comprising the steps of:

s1, image generation, a plurality of images generated by time-series image capture of an object;

s2, determining a main image and a sub image, selecting one image as a main image and a plurality of images as a sub image group according to the static image generation instruction;

s3, extracting contour components from the main image and the sub-image;

s4, judging foreign matters on the lens, judging whether foreign matters exist on the lens according to the relative position of the contour components in the image, if so, executing step S5, if not, determining the main image as a final image and executing step S9;

s5, selecting the peripheral characteristics of the main image foreign matter, and selecting and recording the peripheral characteristics of the main image foreign matter;

s6, extracting images corresponding to the sub-image group, determining the corresponding position of the part of the main image blocked by the foreign matter on the sub-image group according to the peripheral features of the foreign matter of the main image, and extracting the images at the corresponding position;

s7, synthesizing the corresponding images and the main images, splicing the images at the corresponding positions to the foreign matters of the main images, and generating a plurality of synthesized main images;

s8, judging the texture of the synthesized main image, detecting the coincidence degree of the synthesized position of the main image, and selecting one of the synthesized main images with the highest coincidence degree as a final image;

s9, storing and outputting the final image, and storing the multiple synthesized main images, the main images and the sub image groups;

the step of judging the foreign substance on the lens in step S4 includes:

s41, comparing the shape of each contour component of the main image with that of each contour component in the sub image group, and if the shape of the contour component of the main image is the same as that of the contour component in the sub image group, judging that the contour components are the same;

s42, measuring and comparing the lengths of the same contour components from the edges of the main image and the sub-image group on the main image and the sub-image group, if the lengths of the same contour components from the edges of the main image and the sub-image group on the main image and the sub-image group are equal, judging that the same contour components are foreign matters;

the step of discriminating the texture of the synthesized main image in step S8 includes:

s81, re-extracting contour components at the composite splicing position of the composite main image;

s82, performing expansion processing on the contour components at the splicing position of the synthesized main image to generate an expanded synthesized main image;

s83, detecting the goodness of fit of lines at the splicing part of the expanded composite main image;

and S84, comparing the goodness of fit of the multiple expanded composite main images, and selecting the composite main image with the highest goodness of fit as a final image.

2. The endoscope signal processing method according to claim 1, wherein the step of determining the primary and secondary images in step S2 includes:

s21, receiving a static image generation instruction;

s22, receiving the last frame image before the static image generation instruction as the main image;

and S23, taking a plurality of frame images before the last frame image as a secondary image group.

3. The endoscope signal processing method according to claim 2, wherein said step S6 of extracting the image corresponding to the sub-image group comprises:

s61, searching an image area with the peripheral characteristics of the main image foreign matter on the secondary image group;

and S62, intercepting the image area.

4. An endoscope signal processing system, comprising:

an image generation module that generates a plurality of images by time-series imaging of an object;

the main and auxiliary image determining module selects one image as a main image according to the static image generation instruction, and a plurality of images as an auxiliary image group;

a contour extraction module for extracting contour components from the main image and the sub-image;

the foreign matter judging module is used for judging whether foreign matters exist on the lens according to the relative positions of the contour components in the image, if so, the main image foreign matter surrounding characteristic selecting module is activated, and if not, the main image is determined to be a final image and the storage output module is activated;

the main image foreign matter surrounding characteristic selecting module selects and records the characteristics of the main image foreign matter surrounding;

the image extraction module at the corresponding position of the secondary image group determines the corresponding position of the part of the primary image, which is blocked by the foreign matter, on the secondary image group according to the peripheral characteristics of the foreign matter of the primary image, and extracts the image at the corresponding position;

the composite main image module splices the images at the corresponding positions to the foreign matters of the main image to generate a plurality of composite main images;

the texture judging module is used for detecting the goodness of fit of the combined position of the main images and selecting one image with the highest goodness of fit in a plurality of combined main images as a final image;

a storage output module for storing and outputting the final image, and storing the plurality of synthesized main images, the main image, and the sub-image group;

the foreign matter determination module includes:

the contour shape comparison module is used for comparing the shape of each contour component of the main image with that of each contour component in the sub image group, and if the contour component of the main image is the same as the contour component in the sub image group, the main image is judged to be the same contour component;

the outline position comparison module is used for measuring and comparing the lengths of the same outline components from the edges of the main image and the sub-image group on the main image and the sub-image group, and if the lengths of the same outline components from the edges of the main image and the sub-image group on the main image and the sub-image group are equal, the same outline components are judged to be foreign matters;

the texture discriminating module includes:

the contour re-extraction module re-extracts contour components at the synthesized and spliced part of the synthesized main image;

the expansion processing module is used for performing expansion processing on the contour components at the splicing position of the synthesized main image to generate an expanded synthesized main image;

the goodness of fit detection module is used for detecting the goodness of fit of lines at the splicing part of the expanded synthesized main image;

and the goodness of fit contrast module is used for comparing the goodness of fit of the multiple expanded synthesized main images and selecting one synthesized main image with the highest goodness of fit as a final image.

5. The endoscopic signal processing system of claim 4 wherein said primary and secondary image determination module comprises:

an instruction receiving module which receives a static image generation instruction;

a main image selection module for selecting the last frame image before receiving the static image generation instruction as the main image;

and the secondary image selection module selects a plurality of frames of images before the last frame of image as a secondary image group.

6. The endoscopic signal processing system as defined in claim 4, wherein said sub image group corresponding image extracting module comprises:

the comparison characteristic module is used for searching an image area with the peripheral characteristics of the foreign matters of the main image on the auxiliary image group;

and the image intercepting module intercepts the image area.

Images