CN116392077A - A razor back detection system and control method based on three-dimensional depth image - Google Patents

Abstract

The invention relates to a razor back detection and evaluation method based on a three-dimensional depth image, and belongs to the technical fields of photoelectric detection technology and three-dimensional depth image processing. Obtain the three-dimensional depth images of the back of the subject in different poses in the forward bending state through the depth camera; frame the upper and lower edges of the ROI area of the back; identify the feature points of the detection area, and construct the reference line model of the back spine; point cloud data Carry out coordinate conversion and measure the height difference between the left and right sides of the back; combined with the preset razor back height difference threshold, a comprehensive assessment of the subject's back razor back is realized. In the present invention, the three-dimensional depth image obtained through multi-pose acquisition, combined with the digital image processing technology related to artificial intelligence such as feature point extraction and section profile analysis, solves the problem that the traditional razor back detection process relies on the experience of professional doctors and the standard is not uniform , Large human error and large X-ray radiation.

Description

A razor back detection system and control method based on three-dimensional depth image

technical field

The present invention relates to the field of photoelectric detection technology and the field of three-dimensional depth image processing technology, in particular, to a razor back detection system and control method based on a three-dimensional depth image.

Background technique

In my country, scoliosis has become the third major disease that seriously threatens the physical and mental health of children and adolescents after myopia and obesity. In scoliosis, due to the axial rotation of the spine, the ribs on both sides protrude or sag, making the back bilaterally asymmetrical, which not only affects the appearance of the patient's body, but may also cause pain in daily sitting and standing, and poor breathing. Further cause the patient's sense of inferiority to aggravate and a series of psychological barriers. It is a very important examination in scoliosis screening to observe whether there is left and right unevenness in the subject's back through the forward bending test posture, that is, the razor back phenomenon.

The current clinical methods for razor back inspection mainly use methods such as manual visual inspection or manual measurement of the trunk rotation angle with the aid of a scoliosis ruler. According to the diagnostic criteria for patients with scoliosis in the "Clinical Diagnosis and Treatment Guidelines Orthopedics Volume", guide the subjects to perform the forward flexion test. Artificial visual method: the examiner's eyes should be at the same height as the subject's back, and the eyes should be parallel to the subject's bending from head to tail, from the thoracic spine to the lumbar spine, and observe whether the sides of the spine are uneven; use a scoliometer to measure Method: Measure each segment of the subject's spine (thoracic, thoracolumbar, and lumbar) with a trunk rotation measuring instrument, and record the maximum rotation angle and position. If the most serious part of the back asymmetry exceeds 5°, it is highly suspected that the spine Side bend. Both of these methods are based on manual detection, rely heavily on the skills and experience of professional doctors, the standards are not uniform, prone to false detection and missed detection, and the screening effect is not good.

In view of the above problems, in order to meet the scoliosis screening needs of large-scale children and adolescents and realize "early screening, early diagnosis and early treatment", a safe, effective and rapid razor back detection method is urgently needed to realize razor back detection. Automatic collection, detection and evaluation of data.

Contents of the invention

Aiming at the problems of non-uniform razor back inspection standards and complicated procedures in the prior art, the purpose of the present invention is to provide a razor back inspection system and control method based on a three-dimensional depth image.

Embodiments of the invention are achieved through the following technical solutions:

In a first aspect, the present invention provides a method for controlling a razor back detection system based on a three-dimensional depth image, including;

Collect three-dimensional depth images of the subject under different bending angles;

Based on the three-dimensional depth image obtained above, select the range of the upper and lower edges of the back to determine the position of the reference line of the back spine;

Based on the above-mentioned 3D depth image and spine reference line, extract the key points protruding from the 3D depth image along both sides of the reference line, and measure the height difference between the left and right sides of the back;

Analyze and process the height difference data, and comprehensively evaluate the razor back detection results of the subject.

In an embodiment of the present invention, the range of selecting the upper and lower edges of the back also includes;

According to the three-dimensional depth image, determine that the upper edge of the back is the highest position of the back that can be photographed, and the lower edge of the back is the position above the belt of the subject's trousers;

Define the three-dimensional depth data area between the upper and lower edges as the razor back detection area;

In an embodiment of the present invention, the determining the position of the reference line of the back spine further includes;

In the razor back detection area, measure the lowest point of the concave contour of the back section;

Connect all the lowest points and record as where the spine guide is located.

In an embodiment of the present invention, the key points further include;

Along the outline of the subject's back section, with the spine reference line as the center, find the extreme points on the left and right sides that are raised as the key points.

In an embodiment of the present invention, the height difference further includes;

Establish a three-dimensional coordinate system in the XYZ direction;

According to the key point depth information on the left and right sides of the back, calculate the Z-direction data difference, and obtain the current height difference between the left and right sides of the back.

In an embodiment of the present invention, the analyzing and processing the height difference data further includes;

Measure the height difference between the left and right sides of the subject's back along the back from top to bottom;

Extract the data results with the largest difference in height;

Determine that the data result is derived from the data result of the subject's thoracic position, thoracolumbar position or lumbar position;

Process the three-dimensional depth images of the subject's back under different bending angles;

Set the threshold for comparative analysis and get the results.

In the second aspect, a razor back detection system based on a three-dimensional depth image provided by the present invention is characterized in that it includes;

The first module is used to collect and shoot the back shape of the subject under different bending angles when bending forward;

The second module is used for three-dimensional reconstruction of the back and establishment of three-dimensional coordinate points, and then generates a depth model that is highly consistent with the actual shape of the subject's back;

The third module is used to select the ROI area along the upper and lower sides of the subject's back to segment the three-dimensional depth model;

A processing unit, the processing unit is connected to the first module, the second module and the third module, and executes the above-mentioned control method of a razor back detection system based on a three-dimensional depth image.

In a third aspect, an electronic device provided by the present invention includes a memory, a processor, and a computer program stored on the memory and operable on the processor. When the processor executes the computer program, the above-mentioned A control method of razor back detection system based on 3D depth image.

In a fourth aspect, the present invention provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the above-mentioned razor back detection based on a three-dimensional depth image is realized system control method.

The technical solutions of the embodiments of the present invention have at least the following advantages and beneficial effects:

Embodiments of the Invention The present invention focuses on multi-pose acquisition, precise algorithm analysis, efficient data processing, etc., based on three-dimensional depth image data, using artificial intelligence-related digital image processing technologies such as feature point extraction and section contour analysis, to propose a safe , An effective and fast razor back detection method can effectively overcome the traditional razor back inspection process relying on professional physician experience, inconsistent standards, large human errors, large X-ray radiation, and difficulty in large-scale screening.

Description of drawings

In order to illustrate the technical solutions of the embodiments of the present invention more clearly, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention, and thus It should be regarded as a limitation on the scope, and those skilled in the art can also obtain other related drawings based on these drawings without creative work.

Fig. 1 is the general flowchart of detection method of the present invention;

Fig. 2 is a schematic diagram of the subject bending 30 degrees;

Fig. 3 is a schematic diagram of the subject being bent at 45 degrees;

Fig. 4 is a schematic diagram of a subject bent at 90 degrees;

Fig. 5 is a back depth data map of a small angle of forward bending (about 30°);

Fig. 6 is a back depth data map of a large angle of forward bending (about 45°);

Fig. 7 is a back depth data map of forward flexion to the limit (about 90°).

Detailed ways

In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments It is a part of embodiments of the present invention, but not all embodiments. The components of the embodiments of the invention generally described and illustrated in the figures herein may be arranged and designed in a variety of different configurations.

The terms "first, second" and the like in the specification and claims of the present application and the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific sequence or sequence. The naming or numbering of the steps in this application does not mean that the steps in the method flow must be executed in the time/logic order indicated by the naming or numbering, and the steps that have been named or numbered can be implemented according to the needs to be achieved. The execution order of the technical purpose is changed, as long as the same or similar technical effect can be achieved.

The present invention provides a control method of a razor back detection system based on a three-dimensional depth image, the purpose of which is to simplify the traditional razor back detection process and integrate three-dimensional depth image acquisition, artificial intelligence algorithm and razor back evaluation analysis:

Razor back assessment and analysis is to measure the height difference between the left and right sides of the subject's back from top to bottom, extract the data with the largest height difference, and determine that the data comes from the subject's thoracic, thoracolumbar or lumbar section Which segment of the position. By processing the three-dimensional depth images of the subject's back under different bending angles, combined with the comparison and analysis of the set threshold, a comprehensive evaluation result of the razor back inspection is finally generated.

The present invention provides a method for controlling a razor back detection system based on a three-dimensional depth image, comprising;

S101: Collect three-dimensional depth images of the subject under different bending angles;

In this step, it is necessary for the subject to stand in a forward-bending position: the subject’s back to the camera, at a specified distance, feet together, bare back, take a natural standing posture, and complete the forward-bending action according to the prompts of the tester , increasing the angle of the back bend in the forward bend position until a 90° bend is reached.

Collect 3D depth image data: As shown in Figure 2, use the depth camera to collect multiple 3D depth images of the subject's back to analyze whether there is a razor back phenomenon in different parts of the chest, thoracolumbar, and waist.

S102: Based on the three-dimensional depth image obtained above, select the range of the upper and lower edges of the back, and determine the position of the reference line of the back spine;

Carry out target area division on the depth image of the back of the human body acquired in step S101, and generate the direction of the reference line of the back spine.

Target area division: Determine the position of the upper edge of the human body's back and the shooting background, the lower edge of the back, and the subject's trousers belt in the depth image, and delineate the ROI area for razor back detection.

Back spine reference line direction: In the razor back detection area, find the concave lowest point of the vertical back section contour curve, and record it as the location of the spine reference line on the back, repeat this step from top to bottom until the entire razor is generated The spine reference line model of the back detection area, and the direction of the back spine reference line is obtained.

S103: Based on the above-mentioned 3D depth image and the spine reference line, extract key points protruding from the 3D depth image along both sides of the reference line, and measure the height difference between the left and right sides of the back;

According to the reference line of the back spine obtained in step S102, the depth information is used to perform coordinate conversion, to find key points protruding on both sides of the reference line of the back spine, and to measure the height difference between the two sides of the back.

Depth coordinate conversion: use the back spine reference line obtained in step S102 to divide it into multiple segments according to a certain ratio, select the back area for measurement and calculation according to the degree of forward flexion, and use the spine reference line in the segment as the reference line for each posture. As shown in the figure, in the normal vector of the line connecting the first pixel point and the last pixel point of each segment, select a plane together with the Z axis (the direction perpendicular to the camera lens is called the Z axis), which is perpendicular to the XY plane The normal vector is the direction of rotation, and rotate all the 3D point data of this segment around the X-axis by the corresponding angle, and convert all the height of the back of the segment into the new Z-axis coordinates to complete the transformation of the depth coordinates;

Find key points on both sides of the back: As shown in Figures 3, 4, and 5, for different postures corresponding to the thoracic, thoracolumbar, and lumbar segments along the vertical section of the subject’s back, use the spine reference line as a distinction to find the points on both sides. Raise the most obvious key point, so as to locate the 1-3 pixel area where the largest point in the Z direction coordinates is located, which is the coordinate of the key point;

S104: Analyze and process the height difference data, and comprehensively evaluate the razor back detection result of the subject.

Bilateral back height difference: calculate the height difference of the area where the key point is located, first find the maximum height difference of each segment, and then compare the height difference of each segment to find the maximum height difference of the pose.

Integrate the height difference of the bilateral back of the subject in multiple poses, analyze and compare with the set threshold, generate the evaluation conclusion of the razor back, and complete the detection process.

To sum up, the height difference between the left and right sides of the subject's back is measured sequentially along the back from top to bottom, the data with the largest height difference is extracted, and the data is determined to come from the thoracic, thoracolumbar or lumbar section of the subject. Which position. By processing the three-dimensional depth images of the subject's back under different bending angles, combined with the comparison and analysis of the set threshold, a comprehensive evaluation result of the razor back inspection is finally generated.

In an embodiment of the present invention, the selection of the range of the upper and lower edges of the back further includes, according to the three-dimensional depth image, determining that the upper edge of the back is the highest position of the back that can be photographed, and the lower edge of the back is the position above the belt of the subject's trousers , delineate the three-dimensional depth data area between the upper and lower edges as the razor back detection area; in the razor back detection area, measure the lowest point of the concave contour of the back section, connect all the lowest points and record it as the position of the spine reference line.

The present invention also provides a razorback detection system based on a three-dimensional depth image, which is characterized in that it includes;

The first module is used to collect and shoot the back shape of the subject under different bending angles when bending forward;

The second module is used for three-dimensional reconstruction of the back and establishment of three-dimensional coordinate points, and then generates a depth model that is highly consistent with the actual shape of the subject's back;

The third module is used to select the ROI area along the upper and lower sides of the subject's back to segment the three-dimensional depth model;

A processing unit, the processing unit is connected to the first module, the second module and the third module, and executes the above-mentioned control method of a razor back detection system based on a three-dimensional depth image.

A further explanation of the above modules is as follows:

The 3D depth image acquisition mainly uses the depth camera to capture the back shape of the subject under different bending angles when bending forward, realizes the 3D reconstruction of the back and establishes the 3D coordinate points, and then generates a picture that is highly consistent with the actual shape of the subject's back. depth model.

The artificial intelligence algorithm uses a high-performance computer to train and learn the prior knowledge of the back of the human body, automatically selects the ROI area along the upper and lower sides of the subject's back, and divides the three-dimensional depth model; through searching and traversing, the lowest point of the sectional contour depression at different positions on the back is obtained to realize Extraction of the spine reference line on the back of the human body; along the profile of the back section, search for the extremum points on the left and right sides of the spine reference line, and calculate the height difference between the left and right sides of the current position.

In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, each unit may exist separately physically, or two or more units may be integrated into one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.

If the integrated unit is realized in the form of a software function unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. The computer software product is stored in a storage medium, and includes several instructions to enable a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods in various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program codes.

The above are only preferred embodiments of the present invention, and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and changes. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims (9)

1. A razorback detection system control method based on a three-dimensional depth image, characterized in that, comprising; Collect three-dimensional depth images of the subject under different bending angles; Based on the above-mentioned three-dimensional depth image obtained, select the range of the upper and lower edges of the back, and determine the position of the reference line of the back spine; Based on the above-mentioned 3D depth image and the spine reference line, extract the key points protruding from the 3D depth image along both sides of the reference line, and measure the height difference between the left and right sides of the back; Analyze and process the height difference data, and comprehensively evaluate the razor back detection results of the subject. 2. A kind of razor back detection system control method based on three-dimensional depth image according to claim 1, is characterized in that, described selection back upper and lower edge range also comprises; According to the three-dimensional depth image, determine that the upper edge of the back is the highest position of the back that can be photographed, and the lower edge of the back is the position above the belt of the subject's trousers; The three-dimensional depth data area between the upper and lower edges is defined as the razor back detection area. 3. A kind of razor back detection system control method based on three-dimensional depth image according to claim 2, it is characterized in that, said determination back spine reference line location also comprises; In the razor back detection area, measure the lowest point of the concave contour of the back section; Connect all the lowest points and record as where the spine guide is located. 4. A kind of razor back detection system control method based on three-dimensional depth image according to claim 1, is characterized in that, described key point position also comprises; Along the outline of the subject's back section, with the spine reference line as the center, find the extreme points on the left and right sides that are raised as the key points. 5. A kind of razorback detection system control method based on three-dimensional depth image according to claim 4, is characterized in that, described height difference also comprises; Establish a three-dimensional coordinate system in the XYZ direction; According to the key point depth information on the left and right sides of the back, calculate the Z-direction data difference, and obtain the current height difference between the left and right sides of the back. 6. A kind of razor back detection system control method based on three-dimensional depth image according to claim 1, is characterized in that, described height difference data is analyzed and processed and also comprises; Measure the height difference between the left and right sides of the subject's back along the back from top to bottom; Extract the data results with the largest difference in height; Determine that the data result is derived from the data result of the subject's thoracic position, thoracolumbar position or lumbar position; Process the three-dimensional depth images of the subject's back under different bending angles; Set the threshold for comparative analysis and get the results. 7. A razorback detection system based on a three-dimensional depth image, characterized in that, comprising; The first module is used to collect and shoot the back shape of the subject under different bending angles when bending forward; The second module is used for three-dimensional reconstruction of the back and establishment of three-dimensional coordinate points, and then generates a depth model that is highly consistent with the actual shape of the subject's back; The third module is used to select the ROI area along the upper and lower sides of the subject's back to segment the three-dimensional depth model; A processing unit, the processing unit is connected to the first module, the second module and the third module, and executes the control method of a razor back detection system based on a three-dimensional depth image according to any one of claims 1-6 . 8. An electronic device comprising a memory, a processor, and a computer program stored on the memory and operable on the processor, wherein claim 1 is realized when the processor executes the computer program A method for controlling a razor back detection system based on a three-dimensional depth image described in any one of to 6. 9. A computer-readable storage medium, characterized in that a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, a computer program according to any one of claims 1 to 6 is realized. A control method for a razor back detection system based on a three-dimensional depth image.

Images