CN116071337A - Endoscopic image quality evaluation method based on super-pixel segmentation - Google Patents
Endoscopic image quality evaluation method based on super-pixel segmentation Download PDFInfo
- Publication number
- CN116071337A CN116071337A CN202310124389.XA CN202310124389A CN116071337A CN 116071337 A CN116071337 A CN 116071337A CN 202310124389 A CN202310124389 A CN 202310124389A CN 116071337 A CN116071337 A CN 116071337A
- Authority
- CN
- China
- Prior art keywords
- image
- pixel
- super
- quality
- area
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration by the use of local operators
- G06T5/30—Erosion or dilatation, e.g. thinning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/28—Quantising the image, e.g. histogram thresholding for discrimination between background and foreground patterns
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/762—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using clustering, e.g. of similar faces in social networks
- G06V10/763—Non-hierarchical techniques, e.g. based on statistics of modelling distributions
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
-
- 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
- G06T2207/30168—Image quality inspection
Abstract
The application discloses an endoscopic image quality evaluation method based on super-pixel segmentation, which comprises the steps of obtaining an effective area of an endoscopic image through threshold binarization and morphological calculation; secondly, performing super-pixel segmentation on the endoscope image; then, carrying out mean value calculation of HSV color space on each cluster block after super-pixel segmentation; and finally, comparing the average value of each super pixel block with the HSV threshold range of the normal image to obtain the poor-quality area ratio, thereby judging the quality of the endoscope image on the whole. The endoscope image quality evaluation method is consistent with subjective evaluation of doctors, has high reliability, can assist the doctors to screen high-quality endoscope images, and provides assistance for clinical diagnosis of the doctors and artificial intelligent auxiliary diagnosis.
Description
Technical Field
The invention belongs to the technical field of endoscope image quality evaluation, and particularly relates to an endoscope image quality evaluation method based on super-pixel segmentation.
Background
In the endoscope image, because the image itself may contain artifacts such as high light, a large amount of bubble mucus, and the like, and also because of the problem of manual shooting, a large-range light spot or shadow exists in some images, and the movement of the camera may cause motion blur. These artifacts can seriously affect the image quality of the endoscope, which can affect not only the judgment of the doctor on the condition of the patient, but also the artificial intelligence aided diagnosis. High quality endoscopic images are required, whether for medical judgment or scientific research.
Currently, quality evaluation of endoscopic images is mainly performed by subjective evaluation and screening of doctors, a large number of endoscopic images are shot by one patient for each diagnosis, and good pictures in the endoscopic images need to be screened for diagnosis and treatment, so that the method needs very high time consumption. Zhang et al (Zhang T, wang L, gu J, et al design and Implementation ofAReal-time Capsule Endoscope ImageAssessment method. The 3) rd International Conference on Bioinformatics and Biomedical engineering.) it is proposed to evaluate the quality of an endoscopic image without reference by calculating the gradient field by means of the Sobel operator, but the effect is generally because the difference between the texture features of the endoscopic image and the image quality are not highly correlated. Kamen et al (Ali Kamen, shanhui Sun, shaohuaWan, et al Automic tissue differentiationbased on confocal endomicroscopic images for intraoperative guidance in neurosinger. Biomed Res int., 2016.) screen high quality images by calculating image entropy, but such methods have poor experimental results, and the obtained results have a large difference from the subjective judgment of doctors, and cannot effectively help to evaluate image quality.
Disclosure of Invention
The invention aims to solve the problem that a doctor needs to manually screen excellent images from a large number of endoscopic images for diagnosis at present, and provides an endoscopic image quality evaluation method with higher reliability.
In order to achieve the purpose of the invention, the invention discloses an endoscopic image quality evaluation method based on super-pixel segmentation, which is characterized by comprising the following steps:
step 3, counting the mean value of HSV color space channels of the segmented subareas;
and 4, defining an area which is not in a normal threshold range of a preset endoscope image as an inferior area, wherein the inferior area is an inferior picture when the total picture ratio exceeds 25%, and performing quality evaluation on endoscope image data provided by a hospital by using the method, and comparing the result with subjective evaluation of doctors.
Further, step 1 removes the black area in the image by using threshold binarization, and only the effective area and the text description area of the endoscope image are left; creating a circular structural element kernel with the radius of 3, performing morphological processing of corrosion and expansion on the residual image area, removing the text area after the open operation, and only leaving an effective area of the endoscope image, wherein the method specifically comprises the following steps:
step 1-1, performing image threshold binarization processing, setting a pixel value larger than a threshold value 15 to 255, setting a pixel value smaller than the threshold value to 0, and removing black background information of an image;
step 1-2, creating a circular structure kernel B with radius of 3, and convolving the binarized image A with the kernel B
Removing image text information interference by a morphological method;
step 1-3, carrying out morphological corrosion operation on the image: using the anchor point of the kernel B to scan each pixel of the image A, and performing convolution operation on the kernel and the covered binary image, wherein the pixel value is 1 when the result is 1;
step 1-4, performing morphological dilation operation on the image: and (3) scanning each pixel of the image A by using an anchor point of the kernel B, and performing convolution operation on the kernel and the covered binary image, wherein the pixel value is 0 when the result is 0.
Further, in step 2, the endoscopic image is subjected to SLIC super-pixel segmentation, the color image is converted into a 5-dimensional feature vector in CIELAB color space and XY coordinates, then a distance metric is constructed for the 5-dimensional feature vector, the image pixels are locally clustered, and the super-pixels are small areas composed of a series of pixel points that are adjacent in position and have similar color, brightness or texture features.
Further, step 2 specifically includes:
step 2-1, converting the color image into a CIELAB color space and a 5-dimensional feature vector under an XY coordinate system;
step 2-2, clustering according to a distance standard constructed by the 5-dimensional feature vector, and initializing seed points, namely a clustering center: uniformly distributing seed points in the image according to the set number of super pixels;
step 2-3, reselecting the seed points in the n x n field of the seed points;
step 2-4, distributing class labels for each pixel point in the field around each seed point;
step 2-5, performing distance measurement, including color distance and space distance, and respectively calculating the distance between each searched pixel point and the seed point; each pixel point is searched by a plurality of seed points, and the seed point corresponding to the minimum value is taken as the clustering center of the pixel point;
and 2-6, performing iterative optimization, wherein each pixel point is searched by a plurality of seed points, calculating the distance, and selecting the minimum distance as the seed center of the pixel point until the seed center is unchanged.
Further, in step 2, 100 seed points are set, the image has N pixel points in total, K super pixels with the same size are pre-divided, and the distance between adjacent seed points is s=sqrt (N/K);
the distance between each pixel point and the seed point is calculated, and the distance measurement method comprises the following steps:
d c for the color distance d s The space distance is S, the maximum space distance is D, the distance measurement is D, x and y are respectively the abscissa of the pixel point, and l, a and b are pixel color values.
Further, the step 3 specifically includes:
step 3-1, extracting the endoscope image area after the super-pixel segmentation, and converting each super-pixel area into HSV color space which respectively represents hue, saturation and brightness;
step 3-2, setting the threshold ranges of the H, S, V three channels as [5,20], [110,205] and [110,220] respectively through a large number of tests on a high-quality endoscope image dataset (HSV color characteristic histogram statistics is carried out on a large number of endoscope image areas to obtain a good image area HSV color space threshold range which is H [5,20], S [110,205] and V [110,220 ]);
and 3-3, calculating the HSV average value of each super pixel block of the endoscope image to be detected.
Further, in step 3, the method for converting the RGB color space into the HSV color space is as follows:
wherein max is the largest of r, g, b, min is the smallest of r, g, b;
the method for calculating the HSV color space average value of the super pixel block comprises the following steps:
wherein H is AVG Is the hue H mean value S AVG Is the saturation S mean value, V AVG The brightness V-means, and k is the number of pixels.
Further, the step 4 specifically comprises: when the average value of the color space of the super-pixel block HSV of the image to be detected is not in the threshold range of the normal image HSV, the area is considered to be an inferior area; when the average value of the color space of the super-pixel block HSV of the image to be detected is within the threshold range of the normal image HSV, the area is considered to be a normal area; and finally, counting the proportion of the inferior area, and considering the inferior endoscopic image when the proportion is more than 25%.
Further, step 4 specifically includes:
step 4-1, judging that the HSV channel value of the super pixel block exceeds a threshold range as an inferior area, and when the inferior area accounts for more than 25%, recognizing that the image is a low-quality image, so that the diagnosis of doctors and the auxiliary diagnosis of artificial intelligence are greatly influenced;
step 4-2, performing experiments and tests on image data sets of multi-center clinical patients by using an endoscopic image quality evaluation method based on super-pixel segmentation, and comparing experimental results with conventional image quality evaluation results based on Sobel gradient operators, calculation entropy and the like;
and 4-3, comparing the experimental result with subjective evaluation of doctors, and counting the reliability and consistency of the method. Finally, the invention has higher reliability, the prediction result is more consistent with the subjective evaluation of doctors, and the invention is very helpful for the clinical diagnosis of doctors.
Compared with the prior art, the invention has the remarkable progress that: 1) The image quality evaluation method based on super-pixel segmentation is provided, and the problems that a doctor manually screens high-quality images mainly through subjective experience to diagnose the time and labor waste are solved. The method has high reliability, the evaluation result is more consistent with the subjective evaluation of doctors, and can assist the doctors to screen high-quality images for diagnosis and treatment, thereby providing assistance for endoscopy; 2) The HSV color space is used for evaluating the quality of the endoscopic image, and the color features of the endoscopic image are more obvious than the shape and texture features. Compared with RGB color space, HSV color space is closer to human perception, HSV is more sensitive to common brightness distortion of an endoscope image, bubble mucus and other artifacts, and has good recognition on the image artifacts; 3) Compared with common image quality evaluation methods such as gradient, information entropy and full-image color space mean value, the image quality evaluation method has a good image quality evaluation effect, the super-pixel segmentation enables the image to be divided into a series of sub-areas composed of pixel points with similar characteristics, the characteristic statistics of each super-pixel block can better reflect local characteristics of the image, the image quality is reflected from local to whole, and the average characteristic calculation is more accurate and practical than that of the whole image in the past.
In order to more clearly describe the functional characteristics and structural parameters of the present invention, the following description is made with reference to the accompanying drawings and detailed description.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this application, illustrate embodiments of the invention and together with the description serve to explain the invention and do not constitute a limitation on the invention. In the drawings:
FIG. 1 is a general flow chart of an endoscopic image quality evaluation method based on super-pixel segmentation;
FIG. 2 is an endoscopic image preprocessing diagram;
FIG. 3 is a view of a super-pixel segmentation of an endoscopic image;
FIG. 4 is a block diagram of a superpixel partition;
FIG. 5 is a super-pixel partition block HSV diagram;
FIG. 6 is an image raw data and corresponding HSV map;
fig. 7 is a graph of image test results.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all embodiments of the invention; all other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
As shown in fig. 1, in this embodiment, an endoscopic image quality evaluation method based on super-pixel segmentation is disclosed, which includes the following steps:
firstly, image data is read, a pixel value 15 is used as a threshold value binarized image, a circular structural element with the radius of 3 is created as a kernel, an open operation is carried out on a binarized image area, an ROI area is obtained through morphological operation, and a text area in the image is removed.
Then, SLIC super-pixel segmentation is carried out on the preprocessed image, the color image is converted into 5-dimensional feature vectors under CIELAB color space and XY coordinates, then a distance metric is constructed on the 5-dimensional feature vectors, and local clustering is carried out on the image pixels, so that 100 compact and approximately uniform super-pixel areas are generated.
Then, the HSV color space threshold range of the high-quality endoscope image is obtained through experiments, mean value calculation is carried out on HSV color spaces in all super pixel blocks of the image to be detected, the calculated mean value is compared with the HSV threshold of the high-quality endoscope image, and the image occupation ratio of the inferior area is calculated.
And finally, identifying the image to be detected with the inferior area accounting for more than 25% as a low-quality image, comparing the experimental result with the subjective evaluation of a doctor, and finding that the reliability of the quality evaluation result of the endoscope image is high and is consistent with the subjective evaluation of the doctor, so that the doctor can conveniently diagnose and provide an excellent data set for the artificial intelligent auxiliary diagnosis.
Examples
As shown in fig. 2, fig. 2 is a graph of the preprocessing result of the endoscopic image, in which the text region and the black region of the image are removed by thresholding and morphological operation, and the endoscopic image region is retained.
As shown in fig. 3, fig. 3 is a graph of the result of the super-pixel segmentation of the endoscopic image, and 100 super-pixel blocks are generated after the super-pixel segmentation of the endoscopic image, each region has similarity, and the characteristic features are more representative.
As shown in fig. 4, 5 and 6, fig. 4, 5 and 6 are respectively an endoscopic image super-pixel segmentation block diagram, a block HSV diagram and a good image HSV diagram, it can be found that the high-quality and low-quality images have great differences in HSV color space, and after the super-pixel segmentation of the block, whether the average value of the HSV color space in each block is within the normal image threshold value range can be calculated.
As shown in fig. 7, fig. 7 shows quality evaluation results obtained by an endoscopic image quality evaluation method based on super-pixel segmentation, and it can be found that the quality evaluation of an endoscopic image can be well performed by the super-pixel segmentation and the HSV color space, the evaluation results are more consistent with subjective evaluation results of doctors, reliability is greatly improved compared with quality evaluation methods such as Sobel gradient operators and computational entropy, and the like, and the method is suitable for an endoscopic image quality evaluation scene and provides convenience for doctors to diagnose patients.
It is noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.
Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (9)
1. An endoscopic image quality evaluation method based on super-pixel segmentation is characterized by comprising the following steps:
step 1, preprocessing an image shot by an endoscope, and processing the image by a threshold binarization method and a morphological method to obtain a required effective image area;
step 2, performing SLIC super-pixel segmentation on the preprocessed endoscope image by using a linear iterative clustering algorithm;
step 3, counting the mean value of HSV color space channels of the segmented subareas;
and 4, defining an area which is not in a normal threshold range of the preset endoscope image as an inferior area, wherein the inferior area is an inferior picture when the total picture ratio exceeds 25%.
2. The method for evaluating the quality of an endoscopic image based on super-pixel segmentation according to claim 1, wherein step 1 removes a black region in the image by threshold binarization, leaving only an effective region and a text description region of the endoscopic image; creating a circular structural element kernel with the radius of 3, performing morphological processing of corrosion and expansion on the residual image area, removing the text area after the open operation, and only leaving an effective area of the endoscope image, wherein the method specifically comprises the following steps:
step 1-1, performing image threshold binarization processing, setting a pixel value larger than a threshold value 15 to 255, setting a pixel value smaller than the threshold value to 0, and removing black background information of an image;
step 1-2, creating a circular structure kernel B with radius of 3, and convolving the binarized image A with the kernel B
By morphology ofThe method removes the interference of the image text information;
step 1-3, carrying out morphological corrosion operation on the image: using the anchor point of the kernel B to scan each pixel of the image A, and performing convolution operation on the kernel and the covered binary image, wherein the pixel value is 1 when the result is 1;
step 1-4, performing morphological dilation operation on the image: and (3) scanning each pixel of the image A by using an anchor point of the kernel B, and performing convolution operation on the kernel and the covered binary image, wherein the pixel value is 0 when the result is 0.
3. The method for evaluating the quality of an endoscopic image based on superpixel segmentation according to claim 1, wherein in step 2, the endoscopic image is subjected to SLIC superpixel segmentation, a color image is converted into a 5-dimensional feature vector in CIELAB color space and XY coordinates, then a distance metric is constructed for the 5-dimensional feature vector, the image pixels are locally clustered, and the superpixel is a small region composed of a series of pixel points which are adjacent in position and have similar color, brightness or texture features.
4. A method for evaluating the quality of an endoscopic image based on super-pixel segmentation according to claim 1 or 3, wherein the step 2 specifically comprises:
step 2-1, converting the color image into a CIELAB color space and a 5-dimensional feature vector under an XY coordinate system;
step 2-2, clustering according to a distance standard constructed by the 5-dimensional feature vector, and initializing seed points, namely a clustering center: uniformly distributing seed points in the image according to the set number of super pixels;
step 2-3, reselecting the seed points in the n x n field of the seed points;
step 2-4, distributing class labels for each pixel point in the field around each seed point;
step 2-5, performing distance measurement, including color distance and space distance, and respectively calculating the distance between each searched pixel point and the seed point; each pixel point is searched by a plurality of seed points, and the seed point corresponding to the minimum value is taken as the clustering center of the pixel point;
and 2-6, performing iterative optimization, wherein each pixel point is searched by a plurality of seed points, calculating the distance, and selecting the minimum distance as the seed center of the pixel point until the seed center is unchanged.
5. The method for evaluating the quality of an endoscopic image based on superpixel segmentation according to claim 4, wherein in step 2, 100 seed points are set, the image has N pixel points in total, K superpixels with the same size are pre-segmented, and the distance between adjacent seed points is s=sqrt (N/K);
the distance between each pixel point and the seed point is calculated, and the distance measurement method comprises the following steps:
d c for the color distance d s The space distance is S, the maximum space distance is D, the distance measurement is D, x and y are respectively the abscissa of the pixel point, and l, a and b are pixel color values.
6. The method for evaluating the quality of an endoscopic image based on super-pixel segmentation according to claim 1, wherein the step 3 specifically comprises:
step 3-1, extracting the endoscope image area after the super-pixel segmentation, and converting each super-pixel area into HSV color space which respectively represents hue, saturation and brightness;
step 3-2, setting the threshold ranges of the H, S, V three channels to [5,20], [110,205] and [110,220] respectively by a large number of tests on the high quality endoscopic image dataset;
and 3-3, calculating the HSV average value of each super pixel block of the endoscope image to be detected.
7. The method for evaluating the quality of an endoscopic image based on super-pixel segmentation according to claim 6, wherein in step 3, the method for converting the RGB color space into the HSV color space is as follows:
wherein max is the largest of r, g, b, min is the smallest of r, g, b;
the method for calculating the HSV color space average value of the super pixel block comprises the following steps:
wherein H is AVG Is the hue H mean value S AVG Is the saturation S mean value, V AVG The brightness V-means, and k is the number of pixels.
8. The method for evaluating the quality of an endoscopic image based on super-pixel segmentation according to claim 1, wherein the step 4 is specifically: when the average value of the color space of the super-pixel block HSV of the image to be detected is not in the threshold range of the normal image HSV, the area is considered to be an inferior area; when the average value of the color space of the super-pixel block HSV of the image to be detected is within the threshold range of the normal image HSV, the area is considered to be a normal area; and finally, counting the proportion of the inferior area, and considering the inferior endoscopic image when the proportion is more than 25%.
9. The method for evaluating the quality of an endoscopic image based on super-pixel segmentation according to claim 1 or 8, wherein the step 4 specifically comprises:
step 4-1, judging that the HSV channel value of the super pixel block exceeds a threshold range as an inferior region, and when the inferior region accounts for more than 25%, judging that the image is a low-quality image;
step 4-2, performing experiments and tests on image data sets of multi-center clinical patients by using an endoscopic image quality evaluation method based on super-pixel segmentation, and comparing experimental results with conventional image quality evaluation results based on Sobel gradient operators, calculation entropy and the like;
and 4-3, comparing the experimental result with subjective evaluation of doctors, and counting the reliability and consistency of the method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310124389.XA CN116071337A (en) | 2023-02-16 | 2023-02-16 | Endoscopic image quality evaluation method based on super-pixel segmentation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310124389.XA CN116071337A (en) | 2023-02-16 | 2023-02-16 | Endoscopic image quality evaluation method based on super-pixel segmentation |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116071337A true CN116071337A (en) | 2023-05-05 |
Family
ID=86178401
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310124389.XA Pending CN116071337A (en) | 2023-02-16 | 2023-02-16 | Endoscopic image quality evaluation method based on super-pixel segmentation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116071337A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116630311B (en) * | 2023-07-21 | 2023-09-19 | 聊城市瀚格智能科技有限公司 | Pavement damage identification alarm method for highway administration |
-
2023
- 2023-02-16 CN CN202310124389.XA patent/CN116071337A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116630311B (en) * | 2023-07-21 | 2023-09-19 | 聊城市瀚格智能科技有限公司 | Pavement damage identification alarm method for highway administration |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US5452367A (en) | Automated method and system for the segmentation of medical images | |
| US20070206844A1 (en) | Method and apparatus for breast border detection | |
| EP2188779A1 (en) | Extraction method of tongue region using graph-based approach and geometric properties | |
| CN106096491B (en) | Automatic identification method for microaneurysms in fundus color photographic image | |
| CN111223110A (en) | Microscopic image enhancement method and device and computer equipment | |
| CN116071337A (en) | Endoscopic image quality evaluation method based on super-pixel segmentation | |
| Chen et al. | Ulcer detection in wireless capsule endoscopy video | |
| Akshay et al. | Segmentation and classification of FMM compressed retinal images using watershed and canny segmentation and support vector machine | |
| Uribe-Valencia et al. | Automated Optic Disc region location from fundus images: Using local multi-level thresholding, best channel selection, and an Intensity Profile Model | |
| Ali et al. | Medical images enhanced by using fuzzy logic depending on contrast stretch membership function | |
| CN111340773A (en) | Retinal image blood vessel segmentation method | |
| Wisaeng et al. | Automatic detection of exudates in retinal images based on threshold moving average models | |
| Ramasubramanian et al. | A Comprehensive Analysis of Various Delineation method for Exudates in Fundus Images using Miniaturized Pi Board | |
| CN112052854B (en) | Medical image reversible information hiding method for realizing self-adaptive contrast enhancement | |
| Ramella | Automatic Skin Lesion Segmentation based on Saliency and Color. | |
| Badeka et al. | Evaluation of LBP variants in retinal blood vessels segmentation using machine learning | |
| CN111062909A (en) | Method and equipment for judging benign and malignant breast tumor | |
| Arnold et al. | Indistinct frame detection in colonoscopy videos | |
| KR102380560B1 (en) | Corneal Ulcer Region Detection Apparatus Using Image Processing and Method Thereof | |
| CN111652805B (en) | Image preprocessing method for fundus image stitching | |
| CN113222009A (en) | Method for estimating nodule aspect ratio based on thyroid ultrasound video | |
| CN114129188B (en) | Ultrasonic contrast perfusion flow direction estimation method and imaging system | |
| CN117522719B (en) | Bronchoscope image auxiliary optimization system based on machine learning | |
| CN117541800B (en) | Laryngoscope image-based laryngeal anomaly segmentation method | |
| JP4252744B2 (en) | Abnormal shadow detection device |
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 |