A method for identifying endoscopic examination status based on icons and text

Abstract

The present invention provides a method for identifying endoscopic examination status based on icons and texts, which comprises the following steps in sequence: step 101a, obtaining a video frame at the current moment from a video stream; step 102a, calculating and recording the grayscale value of a status area at a fixed position in the video frame; step 103a, comparing the grayscale value of the same area with that of the previous frame; step 104a, performing OCR recognition on the grayscale value change area; step 105a, performing status matching according to the OCR result; the method of the present invention realizes an auxiliary tool through five steps that automatically selects an algorithm model according to the current examination status to perform a specific algorithm task without the intervention of an instrument operator; in clinical work, it can effectively improve the use efficiency of artificial intelligence-assisted diagnosis equipment.

Description

Technical Field

[0001] The present invention relates to the technical field of computers, in particular to video element recognition, and more specifically to a method for recognizing text and icons in an endoscopic video stream. Background Art

[0002] In an AI endoscopy-assisted diagnosis system, there are usually multiple auxiliary diagnosis models for different procedures, different parts, different modes, and different tasks. These models can only be started in a specific mode to obtain correct results. For example, the gastric disease detection model should be started during gastroscopy, the colonoscopy detection model should be started during colonoscopy, and the depth of lesion infiltration should be judged under narrow-band light imaging mode. However, for the diagnostic model, the timing of starting and ending requires manual control, and during the clinical examination, the instrument operator is often unable to operate another computer device.

[0003] In view of this, the inventor specially designed a method for identifying the endoscopic examination status based on icons and texts, and this case was thus generated. Summary of the invention

[0004] In order to solve the above problems, the technical solution of the present invention is as follows:

[0005] A method for identifying an endoscopic examination status based on icons and texts comprises the following steps in sequence:

[0006] Step 101a, obtaining the current video frame from the video stream;

[0007] Step 102a, calculating and recording the grayscale value of the state area at a fixed position in the video frame;

[0008] Step 103a, comparing the grayscale value of the same area with that of the previous frame;

[0009] Step 104a, performing OCR recognition on the gray value change area;

[0010] Step 105a: Perform status matching based on the OCR result.

[0011] Furthermore, in step 101a, the video stream contains a valid signal, and the valid signal should be convertible into RGB three-channel data through the protocol.

[0012] Furthermore, in step 102a, pixel values ​​are obtained according to the set fixed positions, and the corresponding regional grayscale value sum is generated and recorded.

[0013] Furthermore, the set fixed positions include a shutter symbol area, an endoscope type display area, and an electronic staining status display area, which are in the form of graphic symbols and / or character strings respectively.

[0014] Furthermore, the appearance of the shutter symbol indicates that the endoscope has been connected to the endoscope host, otherwise the area will be black and the grayscale value will be close to 0; the endoscope type display area will display the connected endoscope model when the endoscope is connected, including the endoscope type and endoscope model code; the electronic staining status area will display the electronic staining type when the electronic staining function is turned on.

[0015] Furthermore, the endoscope type includes any one of a gastroscope, a colonoscope, a bronchoscope, a nasopharyngeal endoscope, and a magnifying endoscope.

[0016] Furthermore, in step 103a, the grayscale value of each set area of ​​the current frame is compared with the grayscale value of the corresponding area of ​​the previous frame.

[0017] Furthermore, in step 104a, when the grayscale value change of the set area exceeds a threshold, text recognition is performed on the area; the threshold is specifically set to the average value of the two endpoint states, for example, when the text disappears, the average grayscale value is close to 0, and fluctuates by 5 with the noise; when the text appears, the average grayscale value is approximately 120, and fluctuates by 5 with the noise. The threshold can be set to 60.

[0018] Furthermore, in step 105a, when step 104a generates a non-empty OCR recognition result, the result is compared to obtain the current surgical status of the endoscopic examination; and the result comparison is completed through string regularization matching.

[0019] Furthermore, the surgical status of the endoscopic examination refers to the examination type and the electronic staining type.

[0020] The method of the present invention realizes an auxiliary tool through five steps, which does not require the intervention of the instrument operator and automatically selects the algorithm model according to the current examination status to perform a specific algorithm task; in clinical work, it can effectively improve the use efficiency of artificial intelligence-assisted diagnostic equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The drawings described herein are used to provide further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0022] in:

[0023] Figure 1 It is a schematic diagram of the process of the present invention;

[0024] Figure 2 This is an example of the present invention Figure 1 ;

[0025] Figure 3 This is an example of the present invention Figure 2 . DETAILED DESCRIPTION

[0026] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present invention clearer and more understandable, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0027] See also Figures 1 to 3 , is a method for identifying endoscopic examination status based on icons and text as the best embodiment of the present invention, comprising the following steps in sequence:

[0028] Step 101a, obtaining the current video frame from the video stream, wherein the video stream contains a valid signal, and the valid signal should be able to be converted into RGB three-channel data through a protocol, such as conforming to the V4L2 protocol under the Linux system;

[0029] Step 102a, calculate and record the grayscale value of the state area at the fixed position in the video frame. Specifically, the software obtains the pixel value of each fixed position according to the set fixed position, generates the corresponding regional grayscale value sum and records it, and the set fixed position includes the shutter symbol area, the endoscope type display area, and the electronic staining status display area, which are in the form of graphic symbols or / and strings respectively; the appearance of the shutter symbol indicates that the endoscope has been connected to the endoscope host, otherwise the area is black, and the grayscale value is close to 0; the endoscope type display area will display the connected endoscope model when the endoscope is connected, including the endoscope type and the endoscope model code; the electronic staining status area will display the electronic staining type when the electronic staining function is turned on. The endoscope type includes any one of a gastroscope, a colonoscope, a bronchoscope, a nasopharyngeal laryngoscope, and a magnifying endoscope. For example, the endoscope model code "CF-290I" indicates that the colonoscope endoscope model is 290I, and the electronic staining type is such as "NBI", "LCI" and other words. The set position refers to the specific position of the corresponding display area set according to the endoscope brand, which is a built-in parameter of the software.

[0030] Step 103a, comparing the grayscale value of the same area with the previous frame, specifically, comparing the grayscale value of each set area of ​​the current frame with the grayscale value of the corresponding area of ​​the previous frame;

[0031] Step 104a, perform OCR recognition on the grayscale value change area. When the grayscale value change of the set area exceeds the threshold, perform text recognition (optical character recognition, OCR) on the area. OCR refers to the process in which an electronic device checks the characters printed on paper, determines their shapes by detecting dark and light patterns, and then translates the shapes into computer text using character recognition methods; the threshold is specifically set to the average value of the two endpoint states, such as when the text disappears, the average grayscale value is close to 0, and fluctuates by 5 with noise; when the text appears, the average grayscale value is about 120, and fluctuates by 5 with noise. The threshold can be set to 60;

[0032] Step 105a, performs status matching according to the OCR result. Specifically, when the step 104a generates a non-empty OCR recognition result, the result is compared to obtain the surgical status of the current endoscopic examination; and the result comparison is completed through string regularization matching. String regularization matching is a method for determining whether a string contains a certain pattern. In this case, it is used to determine whether the string belongs to a known type classification; the surgical status of the endoscopic examination refers to the examination type and the electronic staining type, where the examination type is such as colonoscopy, gastroscopy, etc.; the electronic staining type is such as NBI, FICE, LCI, BLI, etc.

[0033] Through the above five steps, the endoscopic inspection state recognition method of the present invention can use the acquired current frame to determine the current state of the inspection, and provide the state output to the downstream task for use, so that it can automatically determine which model should be selected to perform which task.

[0034] In summary, the method of the present invention realizes an auxiliary tool through five steps that does not require the intervention of the instrument operator and automatically selects the algorithm model according to the current examination status to perform specific algorithm tasks; in clinical work, it can effectively improve the use efficiency of artificial intelligence-assisted diagnostic equipment.

[0035] The present invention is described above by way of example in conjunction with the accompanying drawings. It is obvious that the specific implementation of the present invention is not limited to the above-mentioned method. As long as various non-substantial improvements are made using the method concept and technical solution of the present invention, or the concept and technical solution of the present invention are directly applied to other occasions without improvement, they are all within the protection scope of the present invention.

Claims

1. A method for identifying endoscopic examination status based on icons and text, characterized in that: The following steps are included in sequence: Step 101a, obtaining the current video frame from the video stream; Step 102a, calculating and recording the grayscale value of the state area at a fixed position in the video frame; Step 103a, comparing the grayscale value of the same area with that of the previous frame; Step 104a, performing OCR recognition on the gray value change area; Step 105a: Perform status matching based on the OCR result.

2. The method for identifying endoscopic examination status based on icons and text according to claim 1, characterized in that: In step 101a, the video stream contains a valid signal, and the valid signal should be convertible into RGB three-channel data through the protocol.

3. The method for identifying endoscopic examination status based on icons and text according to claim 1, characterized in that: In step 102a, pixel values ​​are obtained according to the set fixed positions, and the corresponding regional grayscale value sum is generated and recorded.

4. The method for identifying endoscopic examination status based on icons and text according to claim 3, characterized in that: The set fixed positions include a shutter symbol area, an endoscope type display area, and an electronic staining status display area, which are in the form of graphic symbols and / or character strings.

5. The method for identifying endoscopic examination status based on icons and text according to claim 4, characterized in that: The appearance of the shutter symbol indicates that the endoscope has been connected to the endoscope host, otherwise the area is black and the grayscale value is close to 0; the endoscope type display area will display the connected endoscope model when the endoscope is connected, including the endoscope type and endoscope model code; the electronic staining status area will display the electronic staining type when the electronic staining function is turned on.

6. The method for identifying endoscopic examination status based on icons and text according to claim 5, characterized in that: The endoscope type includes any one of a gastroscope, a colonoscope, a bronchoscope, a nasopharyngeal endoscope, and a magnifying endoscope.

7. The method for identifying endoscopic examination status based on icons and text according to claim 1, characterized in that: In step 103a, the grayscale value of each set area of ​​the current frame is compared with the grayscale value of the corresponding area of ​​the previous frame.

8. The method for identifying endoscopic examination status based on icons and text according to claim 1, characterized in that: In step 104a, when the grayscale value of the set area changes beyond a threshold, text recognition is performed on the area.

9. The method for identifying endoscopic examination status based on icons and text according to claim 8, characterized in that: In step 105a, when step 104a generates a non-empty OCR recognition result, the result is compared to obtain the current surgical status of the endoscopic examination; and the result comparison is completed through string regularization matching.

10. The method for identifying endoscopic examination status based on icons and text according to claim 9, characterized in that: The surgical status of the endoscopic examination refers to the examination type and electronic staining type.

Images