Intelligent counting and marking method of inserting plate type hair follicle hair transplanting device

Abstract

The invention discloses an intelligent counting and marking method of a plugboard type hair follicle hair transplanting device, and relates to the technical field of hair follicle detection and counting, and the method comprises the steps: installing a micropore array plugboard, collecting an image, extracting the characteristics of hair follicle deviation, closure degree and definition of each hole position, and building a hair follicle data set; then, according to the hair density of the patient and transplantation requirements, the distance between the inserting plates is automatically adjusted, and personalized adaptation is completed; then, through multi-angle image acquisition, extracting a hair follicle structure map, a vascular ring feature map and a hair papilla map, evaluating hair follicle activity, calculating a hair follicle activity index, and comparing the hair follicle activity index with a threshold value to determine a hair follicle quality grade; and finally, de-weighting hair follicle entries and generating an intraoperative implantation list to help a doctor to quickly select hair follicles in a hair implantation stage and facilitate post-operation charging. The method can effectively solve the technical problem that manual counting of hair follicles by a hair transplantation doctor is tedious and is easy to make mistakes, and can guarantee the success rate of a hair transplantation operation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of hair follicle detection and counting, in particular to an intelligent counting and marking method for a plug-in hair follicle hair transplantation device. BACKGROUND

[0002] Hair transplantation surgery is a minimally invasive treatment method using autologous hair follicle transplantation. The surgical procedure can generally be divided into two main procedures: hair follicle extraction and hair follicle planting. In clinical practice, doctors often need to extract thousands of hair follicle units from the donor area of the patient within a few hours, and then plant them according to the established plan. During the entire process, the doctor not only needs to maintain a high degree of concentration on the drilling, implantation depth and angle, but also needs to simultaneously complete real-time counting, morphological identification and activity detection of the extracted hair follicles to ensure the postoperative hair follicle survival rate and hair transplantation effect. In addition, hair transplantation surgery is generally charged by the number of implanted hair follicles, and patients are particularly sensitive to the number of implanted hair follicles and the number of surviving hair follicles in the later period. Therefore, there is an urgent need for a method that can quickly count and detect the activity of hair follicles during extraction and implantation, to solve the technical problems of tediousness and error-prone manual counting of hair follicles by hair transplantation doctors, and to ensure the success rate and good evaluation rate of hair transplantation surgery. SUMMARY

[0003] In view of the deficiencies of the prior art, the present application provides an intelligent counting and marking method for a plug-in hair follicle hair transplantation device to solve the problems mentioned in the background art.

[0004] To achieve the above purpose, the present application realizes the following technical scheme: an intelligent counting and marking method for a plug-in hair follicle hair transplantation device, comprising the following steps: Step one, install the plug-in plate with a numbered micropore array on the hair transplantation device platform, and configure the micropore arrangement with an equal interval first initial interval d1, collect the image of the upper surface of the plug-in plate, obtain the original image frame sequence, extract the first feature of the ith hole site, and calculate the relative horizontal offset of the jth hair follicle in the ith hole site and the vertical offset , the jth hair follicle closure index in the ith hole site and the ith hole site image clarity index , and combine to judge the four states of the ith hole site, including; "empty hole", "point hair follicle", "foreign matter" and "damaged hair follicle", and establish a hair follicle data set; Step two, obtain the individual hair density parameters of the patient and estimate the amount of transplant demand, compare the first initial spacing d1 recorded in step one with the effective number of purified hair follicle data set, combine the coverage error between the current arrangement and the target density, automatically calculate and set the second spacing of the micro-hole array of the plug, calculate the matching change rate between the first initial spacing d1 and the second spacing d2 as the array adaptation index, and drive the plug fine tuning mechanism to complete the second spacing setting to realize individualized adaptation; Step three, after completing the individualized adaptation of the double-sided plug, multi-angle and multi-band image acquisition is completed, the hair follicle structure map, blood vessel ring feature map and hair papilla map are extracted respectively, the activity evaluation operation is performed according to the tissue boundary integrity, papilla clarity and blood vessel texture characteristics, the structure clarity coefficient, hair follicle blood supply feature coefficient and tissue integrity coefficient are constructed, and the hair follicle activity index is constructed according to the above, and the hair follicle activity index is compared with the hair follicle activity index threshold TF for comparative analysis, and the hair follicle quality grade is confirmed; Step four, construct the de-duplicated hair follicle item and correspond to the activity index score, automatically generate the intraoperative implantation deployment list, and assist the doctor in quickly selecting hair follicles and counting the number of hair follicles after the operation in the hair follicle planting stage.

[0005] Preferably, step one includes: S11, through the "plug recognition interface" and the "surgery console synchronization module", the two-dimensional code or RFID on the plug is automatically read to generate the unique number PID of the plug; At the same time, the operator's login voucher is read, and the unique number UID of the doctor / nurse and the operation scene number SID of this time are recorded; S12, detect the plug material type, the number of micro-hole array rows and columns, and the first initial spacing d1, and write into the "hair follicle unit original registration list table"; S13, pre-allocate hole number HID(x, y) for each micro-hole in the "hair follicle unit original registration list table", where (x, y) represents the array row and column coordinates; The table fields include: HID(x, y), first initial spacing d1, plug material and registration timestamp; S14, start the "array imaging module", complete panoramic parallel array scanning on the upper surface of the plug under stable lighting conditions, and obtain the original image frame sequence; S15, for the ith hole HID , execute local image cropping from the original image frame sequence, intercept the corresponding single-hole image, and calculate the first feature by using image recognition technology, including: Taking the hole center as the reference, calculate the relative horizontal offset of the jth hair follicle in the ith hole And the vertical offset , the jth hair follicle closure index in the ith hole And the ith hole image clarity index and combined with the judgment of the four states of the ith hole position, including; "empty hole", "point hair follicle", "foreign matter" and "broken hair follicle"; S16, the four states of the above ith hole position and hole position number HID are collectively packaged into "the ith hair follicle unit structured item" and stored in the hair follicle dataset.

[0006] Preferably, the relative horizontal offset of the jth hair follicle in the ith hole position in S15 and the vertical offset , the jth hair follicle closure index in the ith hole position and the ith hole image definition index The steps of calculation include: S151, for the ith hole position, let the center coordinates be , the detection center coordinates of the jth hair follicle in the hole position are , and the relative horizontal offset of the jth hair follicle in the ith hole position is calculated and the vertical offset , in the following way:

[0007] S152, the Canny edge detection algorithm is used to extract the hair follicle edge contour, find the boundary between the hair follicle unit and the background, obtain the edge pixel point set, and the Ramer-Douglas-Peucker algorithm is used to simplify the polygon approximation of the edge pixel point set to obtain the contour point sequence , and the first point and the last point in the contour point sequence are extracted, and the straight line distance between the first point and the last point is calculated , if the contour point sequence is a completely closed polygon, the start point and the end point coincide, and the distance is 0; The sum of the distances between all adjacent points in the contour point sequence is calculated to obtain the total perimeter of the contour line , and the jth hair follicle closure index in the ith hole position is calculated by the following formula : ;

[0008] S153, the single-hole image region intercepted in the ith hole position is subjected to grayscale processing to obtain a grayscale image, the Laplace operator is used to enhance the edge of the grayscale image to obtain a gradient response graph, and the variance of the pixels in the gradient response graph is defined as the ith hole image definition index :​​ ;

[0009] wherein N represents the total number of pixels in the grayscale image, represents the grayscale value of the kth pixel, represents the average grayscale value of all pixels, i.e.: ;

[0010] S154, preset image definition threshold , set to 0.5-0.7; the preset closure threshold , set to 0.7-0.9; set the center offset threshold , allow the hair follicle center to be offset within 10 pixels within a certain range of the hole center; and make the following judgments, including: When , and , judge the ith hole as "empty hole" state; When , and , judge the ith hole as "point hair follicle" state; When the relative horizontal offset of the jth hair follicle in the ith hole and the vertical offset any one of them is greater than , judge that the target of the ith hole is not at the center of the empty set, if at this time, mark the target as "foreign matter" state; On the contrary, the relative horizontal offset of the jth hair follicle in the ith hole and the vertical offset are both less than or equal to , and , then judge as "damaged hair follicle" and record as an abnormal item.

[0011] Preferably, step two includes: S21, by analyzing the scan report of the patient's scalp, the target hair transplantation area Atarget and the current area hair density Dnative are extracted, and according to the patient's postoperative density Dtarget, the dimensionless treatment is calculated to obtain the estimated transplant quantity Nreq, the formula is as follows: ;

[0012] S22, by obtaining the plug geometry parameters, including plug length Lb and plug width Wb, combined with the first initial spacing d1, the dimensionless treatment is calculated to obtain the maximum number of hole sites Ncap under d1, the formula is as follows: ;

[0013] S23. By extracting the number of hair follicles (Navail) from the hair follicle dataset and combining it with the area (Atarget) of the target hair transplant region, after dimensionless processing, the achievable density (Dcurr) of the implant under d1 is calculated using the following formula: ;

[0014] S24. Combining the achievable density Dcurr of the insert under d1 and the patient's postoperative expected density Dtarget, after dimensionless processing, the coverage error rate Ecover is calculated as follows: ;

[0015] S25, Preset coverage error rate threshold Set to 5%; when Ecover≤ If the initial spacing d1 is used, then the calculation of the second spacing d2 and the array adaptation adjustment are performed.

[0016] Preferably, in step S25, the calculation of the second spacing d2 and the array adaptation adjustment are performed, and the specific steps are as follows: S251. Combining the calculated estimated number of transplanted holes Nreq with the calculated maximum number of holes Ncap that can be placed on the insert plate under d1, after dimensionless processing, the second spacing d2 is calculated as follows: ;

[0017] S252. After dimensionless processing based on the difference between d1 and d2, the array fit index Rmatch is calculated using the following formula: ;

[0018] Based on the array adaptation index Rmatch, the drive plate fine-tuning mechanism completes the setting of the second spacing d2 to achieve individualized adaptation.

[0019] Preferably, in step three, multi-angle and multi-band image acquisition is performed on both sides of the individually adapted insert to obtain hair follicle structure map, vascular ring feature map, and hair papilla map; based on tissue boundary integrity, papilla clarity, and vascular texture features, an activity assessment operation is performed to construct the structural clarity coefficient Cstruct, the hair follicle blood supply feature coefficient Cvasc, and the tissue integrity coefficient Cpt; the specific steps include: S311. The construction of the structural clarity coefficient Cstruct is achieved by converting the hair follicle structure map into a grayscale image, applying edge enhancement processing to the horizontal and vertical directions of the image, extracting the gradient changes of the hair follicle edge regions in the image, and statistically analyzing the average intensity of the edge response in the overall image as the edge intensity E of the structure map. S312. Using the gray-level distribution statistical analysis method, extract the standard deviation of the gray-level distribution based on the amplitude of the gray-level change in the image, and use it as the image contrast C. S313. Use the Canny operator to extract the edge contour of the hair follicle structure in the hair follicle structure image to obtain a set of closed-loop boundary lines. Apply the contour tracking algorithm to extract the main boundary of the hair follicle region. Count the breakpoints and successive segments of the main boundary. Compare the number of breakpoints and successive segments with the total length of the boundary to estimate the degree of boundary damage in the image, which is used as the boundary damage ratio R. S314. After dimensionless processing using the edge intensity E of the structural map, image contrast C, and boundary damage ratio R, the structural sharpness coefficient Cstruct is calculated using the following formula: ;

[0020] Preferably, the specific steps for constructing the hair follicle blood supply characteristic coefficient Cvasc are as follows: S321. By standardizing the grayscale of the vascular ring feature map, highlighting the fine vascular structure through the vascular enhancement filter, distinguishing the vascular pixels from the background area using image segmentation technology, counting the number of pixels identified as vascular, and comparing the size of the area containing vascular in the whole image to obtain the vascular density D. S322. Extract the vascular contour by using an edge detection algorithm on the vascular ring feature map, identify all independent vascular structures based on connectivity analysis, count the number of closed-loop structures and the total number of vascular structures, and obtain the vascular closed-loop rate L. S323. By skeletonizing the vascular ring feature map and compressing it into a centerline representation with a single pixel width, all vascular segments are identified, and endpoint pixels in the vascular segment are extracted. The number of vascular breakpoints and the total number of vascular segments are counted to obtain the vascular breakage rate G. S324. Combining vascular density D, vascular closure rate L, and vascular rupture rate G, after dimensionless processing, the hair follicle blood supply characteristic coefficient Cvasc is calculated as follows: .

[0021] Preferably, the specific steps for constructing the organizational integrity coefficient are as follows: S331. By using the Sobel algorithm to perform edge detection on the dermal papilla image, the papilla edges are extracted. Then, through contour tracking and connected component analysis, the edge closure path is identified, and the papilla contour closure degree H is obtained. S332. By performing gray-level co-occurrence matrix analysis on the hair papilla image, local texture features are extracted, the stability of gray-level distribution within the region is analyzed, and blurring, discontinuity, and texture disorder are identified through texture clarity and continuity to obtain internal texture integrity B. S333. The edges of the hair papilla image are extracted using the Sobel algorithm. Then, through morphological processing and contour analysis, edge breakpoints and gaps are identified. The gap area ratio DL is obtained by comparing the gap length with the total contour length. S334. Combining the nipple contour closure H, internal texture integrity B, and notch area ratio DL, after dimensionless processing, the tissue integrity coefficient Cpt is calculated using the following formula: .

[0022] Preferably, step three also includes: S341. Based on the calculated structural clarity coefficient Cstruct, hair follicle blood supply characteristic coefficient Cvasc, and tissue integrity coefficient Cpt, after dimensionless processing, the hair follicle activity index Factive is calculated using the following formula: ;

[0023] In the formula, w1, w2, and w3 represent weighting coefficients; S342. By setting a preset hair follicle activity index threshold TF, and comparing the hair follicle activity index Factive with the hair follicle activity index threshold TF, the following evaluation results are obtained: When the hair follicle activity index Factive is greater than or equal to the hair follicle activity index threshold TF, it is marked as "preferred implantation"; When the hair follicle activity index threshold TF*70%≤hair follicle activity index Factive<hair follicle activity index threshold TF, mark it as "routine implantation"; When the hair follicle activity index Factive is less than the hair follicle activity index threshold TF*70%, it is marked "Use temporarily"; this indicates that the image quality is low or the tissue is severely damaged, and it is recommended to suspend use.

[0024] Preferably, step four includes: S41. For hair follicles that may be repeatedly identified in multi-angle images with the same hole HID, perform fusion and normalization processing based on spatial location and image features to ensure that each hair follicle appears only once in the list, forming the final deduplicated hair follicle entry. S42. The deduplicated hair follicle entries are combined with the hair follicle activity index (Factive) markings of "preferred implantation", "routine implantation" and "temporarily postponed use" to automatically generate an intraoperative implantation and allocation list. This list is displayed and filtered in real time through the doctor's operating terminal, assisting the doctor in quickly selecting hair follicles during the hair follicle transplantation stage and calculating the number of hair follicles after the procedure.

[0025] This invention provides an intelligent counting and marking method for a plug-in type hair follicle transplant device. It has the following beneficial effects: (1) The intelligent counting and marking method of the insert-type hair follicle transplant device, through the insertion plate QR code / RFID identification, full plate imaging and automatic deduplication and normalization processing, the system can complete the real-time counting and classification of thousands of hair follicles at one time, avoiding omissions and duplications in manual statistics; doctors do not need to manually check the quantity during the extraction-planting process which lasts for several hours, thus improving the overall operation efficiency.

[0026] (2) The intelligent counting and marking method of the insert-type hair follicle transplant device relies on the multimodal image structure map, vascular ring and dermal papilla map to extract the three major coefficients of clarity, blood supply and integrity, automatically generate the hair follicle activity index, and use threshold partitioning to mark the "priority / routine / delayed" implantation level; doctors can prioritize the implantation of highly active hair follicles, shorten the time of removal from the body, improve the survival rate of hair follicles and the uniformity of postoperative density, and significantly improve patient satisfaction.

[0027] (3) The intelligent counting and marking method of the insert-type hair follicle transplant device generates an implantation and allocation list in real time according to the activity level and displays it on the terminal. It can be quickly filtered by priority, hole position or quantity. Doctors only need to take it according to the prompts during the implantation stage, reducing the time and effort of frequently checking the hair follicle cup or manually sorting, realizing "seeing is planting", and reducing the risk of long-term surgical fatigue.

[0028] (4) The intelligent counting and marking method of this insert-type hair follicle transplant device automatically records the extraction, activity assessment and implantation status of each hair follicle, generating a complete data chain, since the hair transplant fee is charged based on the "actual number of implanted hair follicles". After the operation, the bill can be quickly reconciled and the survival rate can be tracked, which not only avoids patients' doubts about the number, but also establishes a standardized quality assessment and review system for institutions, improving reputation and compliance. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of the intelligent counting and marking method steps of the insert-type hair follicle transplant device of the present invention. Detailed Implementation

[0030] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] Example 1 Please see Figure 1 This invention provides an intelligent counting and marking method for a plug-in type hair follicle transplant device, comprising the following steps: Step 1: Install the insert plate with a numbered micropore array onto the hair transplant device platform, and configure the micropores with a preset first initial spacing d1. Acquire images of the upper surface of the insert plate to obtain the original image frame sequence, extract the first feature of the i-th pore position, and calculate the relative horizontal offset of the j-th hair follicle in the i-th pore position. and vertical offset The closure index of the j-th hair follicle in the i-th pore. and the image sharpness index of the i-th aperture position In conjunction with the four states of the i-th hole position, including "empty hole", "point hair follicle", "foreign object" and "damaged hair follicle", a hair follicle dataset is established. Step 2: Obtain individual hair density parameters and estimated transplantation needs of the patient. Compare the first initial spacing d1 recorded in Step 1 with the effective number of purified hair follicle datasets. Combine the coverage error between the current arrangement and the target density to automatically calculate and set the second spacing of the micropore array of the insert plate. Calculate the matching change rate between the first initial spacing d1 and the second spacing d2 as the array adaptation index, and drive the insert plate fine-tuning mechanism to complete the second spacing setting to achieve individualized adaptation. Step 3: After acquiring multi-angle and multi-band images of both sides of the individualized fitting insert, extract the hair follicle structure map, vascular ring feature map, and dermal papilla map respectively. Perform activity assessment based on tissue boundary integrity, papilla clarity, and vascular texture features to construct structural clarity coefficient, hair follicle blood supply feature coefficient, and tissue integrity coefficient. Based on these, construct a hair follicle activity index. At the same time, compare and analyze the hair follicle activity index with the hair follicle activity index threshold TF to confirm the hair follicle quality grade. Step 4: Construct duplicate hair follicle entries and match them with the activity index score to automatically generate an intraoperative implantation and allocation list, assisting doctors in quickly selecting hair follicles during the hair follicle transplantation stage and calculating the number of hair follicles after the procedure.

[0032] In this embodiment, through the integrated process of "automatic counting + activity grading + implantation list", doctors do not need to manually count and screen a large number of hair follicles during long and intensive surgery. They can quickly select suitable hair follicles for implantation with the help of real-time prompts generated by the system. This effectively reduces human error and operational burden, shortens the waiting time for hair follicles to leave the body, improves the survival rate of hair follicles and postoperative results, and realizes automatic billing based on the number of implanted follicles, enhancing the efficiency and transparency of the surgery.

[0033] Example 2 This embodiment is an explanation based on Embodiment 1. Please refer to it. Figure 1 Specifically, step one includes: S11. Through the "Plug Identification Interface" and the "Surgical Control Console Synchronization Module", the QR code or RFID on the plug plate is automatically read to generate a unique plug plate number PID; at the same time, the operator's login credentials are read to record the doctor / nurse's unique number UID and the current surgical session number SID. S12. Detect the material type of the insert plate, the number of rows and columns of the numbered micropore array, and the first initial spacing d1, and write them into the "Original Registration List of Hair Follicle Units"; S13. In the "Original Registration List of Hair Follicle Units", pre-assign a hole position number HID(x,y) to each subsequent micropore, where (x,y) represents the array row and column coordinates; the registration table fields include: HID(x,y), first initial spacing d1, insert material and registration timestamp; S14. Start the "array imaging module" to complete a panoramic parallel array scan of the upper surface of the insert plate under stable lighting conditions and obtain the original image frame sequence; S15, HID for the i-th hole position Local image cropping is performed from the original image frame sequence to extract the corresponding single-hole image, and image recognition technology is used to calculate the first feature, including: Using the center of the pore as a reference, calculate the relative horizontal offset of the j-th hair follicle in the i-th pore. and vertical offset The closure index of the j-th hair follicle in the i-th pore. and the image sharpness index of the i-th aperture position It also combines the four states of the i-th hole position to determine: "empty hole", "hair follicle", "foreign object" and "damaged hair follicle". S16. Assign the four states of the i-th hole and the hole number HID. They are collectively encapsulated into a "structured entry for the i-th hair follicle unit" and stored in the hair follicle dataset.

[0034] In this embodiment, through automatic identification, structural registration, and image analysis of the insert plate and micropore array, efficient structured registration and status classification of hair follicle units can be achieved in the preoperative stage. The system automatically generates a unique number and status entry for each pore, ensuring the accuracy and consistency of subsequent hair follicle identification and screening operations, greatly reducing the intensity of manual operation and recording errors, and providing a reliable foundation for rapid intraoperative retrieval and postoperative traceability management.

[0035] Example 3 This embodiment is an explanation based on Embodiment 2. Please refer to it. Figure 1 Specifically, the relative horizontal offset of the j-th hair follicle in the i-th pore in S15 and vertical offset The closure index of the j-th hair follicle in the i-th pore. and the image sharpness index of the i-th aperture position The calculation steps include: S151. For the i-th hole, let its center coordinates be... With the center coordinates as Based on the reference, the coordinates of the detection center of the j-th hair follicle in this pore position are: Then calculate the relative horizontal offset of the j-th hair follicle in the i-th hole. and vertical offset The method is as follows: ;

[0036] The relative horizontal offset of the j-th hair follicle in the i-th aperture. and vertical offset In order to accurately determine the specific location of each hair follicle within the pore, so as to facilitate subsequent positioning, statistics and adjustment; S152. The Canny edge detection algorithm is used to extract the hair follicle edge contour, find the boundary between the hair follicle unit and the background, and obtain the edge pixel set. The Ramer-Douglas-Peucker algorithm is used to approximate and simplify the edge pixel set into polygons, and obtain the contour point sequence. And extract the contour point sequence. Given the first and last points, calculate the straight-line distance between them. If the contour point sequence If the polygon is completely closed, then the starting point and ending point coincide, and the distance between them is 0. Calculate the contour point sequence The total perimeter of the outline is obtained by summing the distances between all adjacent points. The closure index of the j-th hair follicle in the i-th pore is calculated using the following formula. : ;

[0037] If the distance between the start and end points is very small (almost closed), the closure degree is close to 1, indicating that the contour is well closed; if the distance is large, it means that the contour is open, the closure degree is less than 1, and the contour has a risk of gaps or breaks. S153. The single-hole image region captured from the i-th hole position is converted to grayscale to obtain a grayscale image. The Laplacian operator is used to enhance the edges of the grayscale image to obtain a gradient response image. The variance of the pixels in the gradient response image is defined as the image sharpness index of the i-th hole position. : ;

[0038] Where N represents the total number of pixels in the grayscale image. This represents the grayscale value of the k-th pixel. This represents the average grayscale value of all pixels, i.e.: ;

[0039] The larger the value, the richer the image details and the higher the clarity; if the image is blurry, the edge response is weak, the pixel changes are small, and the variance approaches zero. S154, Preset image sharpness threshold The threshold value was set to 0.5-0.7. Based on extensive experimental data and image processing results, combined with expert experience and standard image quality requirements, the distribution range of sharpness coefficients for sharp and blurry images was statistically determined by measuring the edge response, texture detail, and sharpness of various image types at different resolutions. Referring to standard specifications and image analysis accuracy requirements in the field of image processing, a suitable sharpness threshold range was determined. This threshold setting is used to accurately determine image sharpness, ensuring that the hair follicle image quality meets the requirements for surgical operation and avoiding low-quality images affecting hair follicle recognition and localization. Preset closure threshold The closure value is set to 0.7-0.9. Based on a large amount of hair follicle image edge detection and contour analysis data, a reasonable closure range is set by statistically analyzing the proportion of well-closed and gapped hair follicle boundaries, combined with expert experience and image feature analysis. According to the integrity of the hair follicle contour and image processing requirements, and referring to image analysis standards and the closure characteristics of hair follicle histology, a closure threshold is determined. This threshold is used to assess the integrity of the hair follicle structure, ensuring that the boundaries of the hair follicle image are free from obvious damage or loss, thus guaranteeing the effectiveness of the hair follicle. Set center offset threshold The procedure allows the hair follicle center to deviate by no more than 10 pixels from the center of the pore. Based on the analysis of offset data from multiple sets of hair follicle images, a reasonable offset range is derived by statistically analyzing the positions of the hair follicle center and the pore center, combined with expert experience and the precision requirements of hair follicle transplantation. Combining the actual surgical precision requirements and offset tolerance, and referring to practical experience in hair follicle transplantation and the precision requirements of machine vision algorithms, an offset threshold is determined. This threshold ensures that the hair follicle can be accurately positioned at the center of the pore, preventing positional deviations from affecting the hair transplantation effect. And make the following judgments, including: when ,and When the i-th hole is determined to be in an "empty hole" state, it indicates that there is no hair follicle structure and no significant edge; the image has almost no texture (blurred or pure background) and no closed contour can be detected → indicating that there is no hair follicle tissue in the hole; when ,and When the i-th hole is identified as a "dot hair follicle" state, it indicates the presence of a clear single hair follicle structure; the image is clear, the outline is complete and closed, and the target is basically located at the geometric center of the hole → a complete hair follicle; When the relative horizontal offset of the j-th hair follicle in the i-th hole and vertical offset Any item > Determine if the target of the i-th hole is not located at the center of the empty space set. If the target is marked as a "foreign object", then the target is marked as a "foreign object" state; otherwise, the relative horizontal offset of the j-th hair follicle in the i-th hole is... and vertical offset All ≤ ,and If the condition is found to be "damaged hair follicle" and recorded as an abnormal entry, it will be identified as such.

[0040] The specific principle behind the "foreign object" state is: a clear image indicates the presence of an identifiable target, but the target is off-center, mostly foreign objects such as instrument debris or cotton fibers being adsorbed to the edge of the hole. The specific principle behind the "damaged hair follicle" state is: the target is still centered, but the outline is missing / broken, and the closure is low → the hair follicle is damaged or incomplete.

[0041] In this embodiment, by automatically calculating the hair follicle center offset, contour closure and image clarity within a single-hole image, and accordingly determining the four states of "empty hole, single hair follicle, foreign body, and broken hair follicle", the system can screen out damaged or foreign body-containing holes during the extraction stage, ensuring that only healthy hair follicles with a centered position and complete structure enter the subsequent implantation process. This significantly reduces the pressure of manual verification during the operation, reduces the risk of foreign body contamination, and improves the overall quality of the final implanted hair follicles.

[0042] Example 4 This embodiment is an explanation based on Embodiment 1. Please refer to it. Figure 1 Specifically, step two includes: S21. By analyzing the patient's scalp scan report, extract the target hair transplant area Atarget and the current hair density Dnative. Based on the patient's postoperative expected density Dtarget, after dimensionless processing, calculate the estimated number of transplants Nreq, as shown in the following formula: ;

[0043] S22. By obtaining the geometric parameters of the insert plate, including the insert plate length Lb and the insert plate width Wb, and combining them with the first initial spacing d1, after dimensionless processing, the maximum number of holes Ncap that can be placed in the insert plate under d1 is calculated, as follows: ;

[0044] S23. By extracting the number of hair follicles (Navail) from the hair follicle dataset and combining it with the area (Atarget) of the target hair transplant region, after dimensionless processing, the achievable density (Dcurr) of the implant under d1 is calculated using the following formula: ;

[0045] S24. Combining the achievable density Dcurr of the insert under d1 and the patient's postoperative expected density Dtarget, after dimensionless processing, the coverage error rate Ecover is calculated as follows: ;

[0046] S25, Preset coverage error rate threshold The threshold was set at 5%. Based on statistical analysis of a large amount of hair transplant surgery data and the density distribution of the transplanted area, and combined with the difference between the actual postoperative hair follicle coverage effect and the patient's expected density, the range of variation of the coverage error rate was extracted. By combining the experience of industry experts and referring to the tolerance standards for coverage error in hair transplant surgery, a reasonable error rate threshold was set. This threshold was established by referring to hair transplant industry standards, minimum coverage error rate requirements provided by equipment manufacturers, and clinical experience to effectively identify and adjust the density coverage of the transplanted area, ensuring that the hair transplant surgery achieves the expected results and avoiding excessive or insufficient hair transplant density from affecting the postoperative outcome. When Ecover≤ If the initial spacing d1 is used, then the calculation of the second spacing d2 and the array adaptation adjustment are performed.

[0047] In this embodiment, by comparing the patient's postoperative desired density with the actual achievable density of the inserts at the current spacing in a dimensionless manner, the coverage error rate is calculated, and it is determined whether the spacing of the inserts needs to be adjusted accordingly. This enables individualized dynamic adaptation of the hair transplant array, ensuring not only the matching accuracy of the transplant density per unit area but also avoiding resource waste and insufficient planting density, thereby improving the postoperative aesthetic effect and the success rate of hair transplantation.

[0048] Example 5 This embodiment is an explanation based on Embodiment 4. Please refer to it. Figure 1 Specifically, in step S25, the calculation of the second spacing d2 and the array adaptation adjustment are performed. The specific steps are as follows: S251. Combining the calculated estimated number of transplanted holes Nreq with the calculated maximum number of holes Ncap that can be placed on the insert plate under d1, after dimensionless processing, the second spacing d2 is calculated as follows: ;

[0049] S252. After dimensionless processing based on the difference between d1 and d2, the array fit index Rmatch is calculated using the following formula: ;

[0050] Based on the array adaptation index Rmatch, the drive plate fine-tuning mechanism completes the setting of the second spacing d2 to achieve individualized adaptation.

[0051] In this embodiment, by introducing a second spacing d2 and an array adaptation index Rmatch, the spacing of the micro-holes in the insert plate can be dynamically calculated and adjusted when the estimated number of transplants does not match the initial hole capacity. This drives the insert plate fine-tuning mechanism to achieve precise spacing reconstruction, ensuring that the number of holes is highly matched with the individual needs of the patient, and improving the customization and operational efficiency of the hair transplant plan.

[0052] Example 6 This embodiment is an explanation based on Embodiment 5. Please refer to it. Figure 1 Specifically, in step three, multi-angle and multi-band image acquisition is performed on both sides of the individually adapted insert to obtain hair follicle structure maps, vascular ring feature maps, and dermal papilla maps; based on tissue boundary integrity, papilla clarity, and vascular texture features, an activity assessment operation is performed to construct the structural clarity coefficient Cstruct, the hair follicle blood supply feature coefficient Cvasc, and the tissue integrity coefficient Cpt; the specific steps include: S311. The construction of the structural clarity coefficient Cstruct is achieved by converting the hair follicle structure map into a grayscale image, applying edge enhancement processing to the horizontal and vertical directions of the image, extracting the gradient changes of the hair follicle edge regions in the image, and statistically analyzing the average intensity of the edge response in the overall image as the edge intensity E of the structure map. S312. Using the gray-level distribution statistical analysis method, extract the standard deviation of the gray-level distribution based on the amplitude of the gray-level change in the image, and use it as the image contrast C. S313. Use the Canny operator to extract the edge contour of the hair follicle structure in the hair follicle structure image to obtain a set of closed-loop boundary lines. Apply the contour tracking algorithm to extract the main boundary of the hair follicle region. Count the number of breakpoints and successive segments on the main boundary. Compare the number of breakpoints and successive segments with the total length of the boundary to estimate the degree of boundary damage in the image, which is used as the boundary damage ratio R. This is used to reflect whether there are any problems such as damage, breakage or blurring of the boundary in the hair follicle structure image. S314. After dimensionless processing using the edge intensity E of the structural map, image contrast C, and boundary damage ratio R, the structural sharpness coefficient Cstruct is calculated using the following formula: ;

[0053] In this embodiment, by acquiring images from multiple angles and in multiple bands and constructing the structural clarity coefficient (Cstruct), multiple image quality factors such as edge strength, grayscale contrast, and boundary continuity can be integrated to effectively identify blurring, breakage, and edge defects in the hair follicle structure image. This enables a quantitative assessment of the integrity of the hair follicle structure and the imaging quality, thereby providing a reliable basis for subsequent activity determination and hair follicle selection, and improving the accuracy and diagnostic value of image analysis.

[0054] Example 7 This embodiment is an explanation based on Embodiment 6. Please refer to it. Figure 1 Specifically, the steps for constructing the hair follicle blood supply characteristic coefficient Cvasc are as follows: S321. By standardizing the grayscale of the vascular ring feature map, highlighting the fine vascular structure through the vascular enhancement filter, distinguishing the vascular pixels from the background area using image segmentation technology, counting the number of pixels identified as vascular, and comparing the size of the area containing vascular in the whole image to obtain the vascular density D. S322. Extract the vascular contour by using an edge detection algorithm on the vascular ring feature map, identify all independent vascular structures based on connectivity analysis, count the number of closed-loop structures and the total number of vascular structures, and obtain the vascular closed-loop rate L. S323. By skeletonizing the vascular ring feature map and compressing it into a centerline representation with a single pixel width, all vascular segments are identified, and endpoint pixels in the vascular segment are extracted. The number of vascular breakpoints and the total number of vascular segments are counted to obtain the vascular breakage rate G. S324. Combining vascular density D, vascular closure rate L, and vascular rupture rate G, after dimensionless processing, the hair follicle blood supply characteristic coefficient Cvasc is calculated as follows: ;

[0055] In this embodiment, by constructing the hair follicle blood supply characteristic coefficient Cvasc and integrating three types of indicators—vascular density, closure rate, and breakage rate—a comprehensive quantitative assessment of the continuity, integrity, and blood supply potential of microvessel distribution in the vascular ring characteristic map can be performed. This allows for accurate identification of the health status of the hair follicle blood supply structure, effective screening of hair follicle units with insufficient blood supply or abnormal vascular structures, and provides an important basis for predicting the activity and optimal distribution of transplanted hair follicles, thereby improving postoperative survival rate and overall efficacy.

[0056] Example 8 This embodiment is an explanation based on Embodiment 6. Please refer to it. Figure 1 Specifically, the steps for constructing the organizational integrity coefficient are as follows: S331. By using the Sobel algorithm to perform edge detection on the dermal papilla image, the papilla edges are extracted. Then, through contour tracking and connected component analysis, the edge closure path is identified, and the papilla contour closure degree H is obtained. S332. By performing gray-level co-occurrence matrix analysis on the hair papilla image, local texture features are extracted, the stability of gray-level distribution within the region is analyzed, and blurring, discontinuity, and texture disorder are identified through texture clarity and continuity to obtain internal texture integrity B. S333. The edges of the hair papilla image are extracted using the Sobel algorithm. Then, through morphological processing and contour analysis, edge breakpoints and gaps are identified. The gap area ratio DL is obtained by comparing the gap length with the total contour length. S334. Combining the nipple contour closure H, internal texture integrity B, and notch area ratio DL, after dimensionless processing, the tissue integrity coefficient Cpt is calculated using the following formula: ;

[0057] In this embodiment, by simultaneously quantifying the closure degree of the papilla contour, the integrity of the internal texture, and the proportion of boundary gaps in the dermal papilla image and synthesizing them into a tissue integrity coefficient Cpt, the system can automatically identify low-quality hair follicles with structural defects or disordered textures, eliminate potentially damaged units in advance, and ensure that the hair follicles entering the implantation stage have complete and healthy papilla tissue, thereby improving the physiological activity and postoperative survival rate of transplanted hair follicles from the source.

[0058] Example 9 This embodiment is an explanation based on Embodiment 8. Please refer to it. Figure 1 Specifically, step three also includes: S341. Based on the calculated structural clarity coefficient Cstruct, hair follicle blood supply characteristic coefficient Cvasc, and tissue integrity coefficient Cpt, after dimensionless processing, the hair follicle activity index Factive is calculated using the following formula: ;

[0059] In the formula, w1, w2, and w3 represent weighting coefficients. The weighting coefficients w1, w2, and w3 are set based on statistical analysis of a large amount of hair transplant surgery data, considering the different importance of hair follicle structure, blood supply characteristics, and tissue integrity on the hair follicle activity index. By calculating the activity index under different hair follicle states, the contribution of each indicator to the assessment of hair follicle activity is analyzed. Combined with expert experience and clinical needs, the relative weight of each coefficient is determined. Generally, all weighting coefficients can be initially set to equal values ​​w1=w2=w3=1. However, in practical applications, the weights may be adjusted according to different patient conditions and hair transplant goals to more accurately reflect the impact of each characteristic on activity assessment. Reasonable setting of weighting coefficients ensures that the hair follicle activity index accurately reflects the actual state of the hair follicles, providing a scientific basis for personalized adjustments to the surgical plan. S342. By setting a preset hair follicle activity index threshold TF, and comparing the hair follicle activity index Factive with the hair follicle activity index threshold TF, the following evaluation results are obtained: When the hair follicle activity index (Factive) is greater than or equal to the hair follicle activity index threshold (TF), it is marked as "preferred for implantation"; this indicates that the hair follicle structure is clear, the blood supply is good, and the tissue integrity is high, and it is recommended to use it first. When the hair follicle activity index threshold TF*70%≤hair follicle activity index Factive<hair follicle activity index threshold TF, it is marked as "routine implantation"; indicating that the imaging structure is average, the blood supply and tissue condition are acceptable, and it can be used routinely. When the hair follicle activity index Factive is less than the hair follicle activity index threshold TF*70%, it is marked "Use temporarily"; this indicates that the image quality is low or the tissue is severely damaged, and it is recommended to suspend use.

[0060] The setting of the hair follicle activity index (TF) threshold relies on statistical analysis of a large amount of hair transplant surgery data, combined with feedback data on the actual hair follicle survival rate and hair transplant effect after surgery. By analyzing the postoperative success rate and hair follicle survival corresponding to different activity index values, and combining hair transplant industry standards and expert experience, a reasonable hair follicle activity index threshold is determined. Generally, the hair follicle activity index threshold is set as a standard value of the hair follicle activity index (Factive) to distinguish the decision criteria for preferred implantation, conventional implantation, and temporary postponement of different hair follicles. By dynamically adjusting this threshold, the hair transplant plan can be optimized, the hair follicle survival rate can be maximized, the occurrence of postoperative complications can be reduced, and the overall surgical effect can be improved.

[0061] In this embodiment, by comprehensively considering the structural clarity, blood supply characteristics, and tissue integrity of hair follicles, the calculated hair follicle activity index can accurately assess the quality of hair follicles and automatically classify them into "preferred implantation," "routine implantation," or "temporarily deferred use" based on a set threshold. This precise classification method helps to efficiently screen hair follicles during the procedure, ensuring that high-quality hair follicles are used preferentially, reducing the postoperative failure rate, and improving the postoperative hair growth effect for patients.

[0062] Example 10 This embodiment is an explanation based on Embodiment 9. Please refer to it. Figure 1 Specifically, step four includes: S41. For hair follicles that may be repeatedly identified in multi-angle images with the same hole HID, perform fusion and normalization processing based on spatial location and image features to ensure that each hair follicle appears only once in the list, forming the final deduplicated hair follicle entry. S42. The deduplicated hair follicle entries are combined with the hair follicle activity index (Factive) markings of "preferred implantation", "routine implantation" and "temporarily postponed use" to automatically generate an intraoperative implantation and allocation list. This list is displayed and filtered in real time through the doctor's operating terminal, assisting the doctor in quickly selecting hair follicles during the hair follicle transplantation stage and calculating the number of hair follicles after the procedure.

[0063] In this embodiment, by fusing spatial location and image features of repeatedly identified hair follicles in the same hole, each hair follicle is recorded only once, thus avoiding data redundancy. Combining the hair follicle activity index and implantation markers, the automatically generated intraoperative implantation allocation list helps doctors efficiently and accurately select suitable hair follicles, greatly improving the operational efficiency of the hair transplant process and providing a precise basis for postoperative billing based on the number of hair follicles, ensuring the personalization and precision of the surgery.

[0064] The threshold is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by those skilled in the art for each set of sample data; as long as it does not affect the ratio between the parameter and the quantized value, it is acceptable.

[0065] The above formulas are all derived from software simulation using a large amount of data, and are selected to be close to the actual values. The coefficients in the formulas are set by those skilled in the art based on the actual situation. The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or changes made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for intelligent counting and marking of insert-type hair follicle transplantation devices, characterized in that, Includes the following steps: Step 1: Install the insert plate with a numbered micropore array onto the hair transplant device platform, and configure the micropores with a preset first initial spacing d1. Acquire images of the upper surface of the insert plate to obtain the original image frame sequence, extract the first feature of the i-th pore position, and calculate the relative horizontal offset of the j-th hair follicle in the i-th pore position. and vertical offset The closure index of the j-th hair follicle in the i-th pore. and the image sharpness index of the i-th aperture position In conjunction with the four states of the i-th hole position, including "empty hole", "hair follicle", "foreign object" and "damaged hair follicle", a hair follicle dataset is established. Step 2: Obtain individual hair density parameters and estimated transplantation needs of the patient. Compare the first initial spacing d1 recorded in Step 1 with the effective number of purified hair follicle datasets. Combine the coverage error between the current arrangement and the target density to automatically calculate and set the second spacing of the micropore array of the insert plate. Calculate the matching change rate between the first initial spacing d1 and the second spacing d2 as the array adaptation index, and drive the insert plate fine-tuning mechanism to complete the second spacing setting to achieve individualized adaptation. Step 3: After acquiring multi-angle and multi-band images of both sides of the individualized fitting insert, extract the hair follicle structure map, vascular ring feature map, and dermal papilla map respectively. Perform activity assessment based on tissue boundary integrity, papilla clarity, and vascular texture features to construct structural clarity coefficient, hair follicle blood supply feature coefficient, and tissue integrity coefficient. Based on these, construct a hair follicle activity index. At the same time, compare and analyze the hair follicle activity index with the hair follicle activity index threshold TF to confirm the hair follicle quality grade. Step 4: Construct duplicate hair follicle entries and match them with the activity index score to automatically generate an intraoperative implantation and allocation list, assisting doctors in quickly selecting hair follicles during the hair follicle transplantation stage and calculating the number of hair follicles after the procedure.

2. The intelligent counting and marking method for the insert-type hair follicle transplant device according to claim 1, characterized in that, Step one includes: S11. Through the "Plug Identification Interface" and the "Surgical Control Console Synchronization Module", the QR code or RFID on the plug plate is automatically read to generate a unique plug plate number PID; at the same time, the operator's login credentials are read to record the doctor / nurse's unique number UID and the current surgical session number SID. S12. Detect the material type of the insert plate, the number of rows and columns of the numbered micropore array, and the first initial spacing d1, and write it into the "Original Registration List of Hair Follicle Units"; S13. In the "Original Registration List of Hair Follicle Units", pre-assign a hole number HID(x,y) to each subsequent micropore, where (x,y) represents the array row and column coordinates; the registration table fields include: HID(x,y), first initial spacing d1, insert material and registration timestamp; S14. Start the "array imaging module" to complete a panoramic parallel array scan of the upper surface of the insert plate under stable lighting conditions and obtain the original image frame sequence; S15, HID for the i-th hole position Local image cropping is performed from the original image frame sequence to extract the corresponding single-hole image, and image recognition technology is used to calculate the first feature, including: Using the center of the pore as a reference, calculate the relative horizontal offset of the j-th hair follicle in the i-th pore. and vertical offset The closure index of the j-th hair follicle in the i-th pore. and the image sharpness index of the i-th aperture position It also combines the four states of the i-th hole position to determine: "empty hole", "hair follicle", "foreign object" and "damaged hair follicle". S16. Assign the four states of the i-th hole and the hole number HID. They are collectively encapsulated into a "structured entry for the i-th hair follicle unit" and stored in the hair follicle dataset.

3. The intelligent counting and marking method for the insert-type hair follicle transplant device according to claim 2, characterized in that, The relative horizontal offset of the j-th hair follicle in the i-th aperture in S15 and vertical offset The closure index of the j-th hair follicle in the i-th pore. and the image sharpness index of the i-th aperture position The calculation steps include: S151. For the i-th hole, let its center coordinates be... With the center coordinates as Based on the reference, the coordinates of the detection center of the j-th hair follicle in this pore position are: Then calculate the relative horizontal offset of the j-th hair follicle in the i-th hole. and vertical offset The method is as follows: ； S152. The Canny edge detection algorithm is used to extract the hair follicle edge contour, find the boundary between the hair follicle unit and the background, and obtain the edge pixel set. The Ramer-Douglas-Peucker algorithm is used to approximate and simplify the edge pixel set into polygons, and obtain the contour point sequence. Extract the contour point sequence. Given the first and last points, calculate the straight-line distance between them. If the contour point sequence If the polygon is completely closed, then the starting point and ending point coincide, and the distance between them is 0. Calculate the contour point sequence The total perimeter of the outline is obtained by summing the distances between all adjacent points. The closure index of the j-th hair follicle in the i-th pore is calculated using the following formula. : ； S153. The single-hole image region captured from the i-th hole position is converted to grayscale to obtain a grayscale image. The Laplacian operator is used to enhance the edges of the grayscale image to obtain a gradient response image. The variance of the pixels in the gradient response image is defined as the image sharpness index of the i-th hole position. : ； Where N represents the total number of pixels in the grayscale image. This represents the grayscale value of the k-th pixel. This represents the average grayscale value of all pixels, i.e.: ; S154, Preset image sharpness threshold Set to 0.5-0.7; preset closure threshold. Set it to 0.7-0.9; Set center offset threshold It allows the hair follicle center to be offset within 10 pixels from the center of the pore within a certain range; and performs the following judgments, including: when ,and When the i-th hole position is determined, it is in the "empty hole" state; when ,and When the i-th hole is determined, it is in the "follicle point" state; When the relative horizontal offset of the j-th hair follicle in the i-th hole and vertical offset Any item > Determine if the target of the i-th hole is not located at the center of the empty space set. If so, the target is marked as a "foreign object"; Conversely, the relative horizontal offset of the j-th hair follicle in the i-th pore position and vertical offset All ≤ ,and If the condition is found to be "damaged hair follicle" and recorded as an abnormal entry, it will be identified as such.

4. The intelligent counting and marking method for the insert-type hair follicle transplant device according to claim 1, characterized in that, Step two includes: S21. By analyzing the patient's scalp scan report, extract the target hair transplant area Atarget and the current hair density Dnative. Based on the patient's postoperative expected density Dtarget, after dimensionless processing, calculate the estimated number of transplants Nreq, as shown in the following formula: ； S22. By obtaining the geometric parameters of the insert plate, including the insert plate length Lb and the insert plate width Wb, and combining them with the first initial spacing d1, after dimensionless processing, the maximum number of holes Ncap that can be placed in the insert plate under d1 is calculated, as follows: ； S23. By extracting the number of hair follicles (Navail) from the hair follicle dataset and combining it with the area (Atarget) of the target hair transplant region, after dimensionless processing, the achievable density (Dcurr) of the implant under d1 is calculated using the following formula: ； S24. Combining the achievable density Dcurr of the insert under d1 and the patient's postoperative expected density Dtarget, after dimensionless processing, the coverage error rate Ecover is calculated as follows: ； S25, Preset coverage error rate threshold Set to 5%; when Ecover≤ If the initial spacing d1 is used, then the calculation of the second spacing d2 and the array adaptation adjustment are performed.

5. The intelligent counting and marking method for the insert-type hair follicle transplant device according to claim 4, characterized in that, In step S25, the calculation of the second spacing d2 and the array adaptation adjustment are performed. The specific steps are as follows: S251. Combining the calculated estimated number of transplanted holes Nreq with the calculated maximum number of holes Ncap that can be placed on the insert plate under d1, after dimensionless processing, the second spacing d2 is calculated as follows: ； S252. After dimensionless processing based on the difference between d1 and d2, the array fit index Rmatch is calculated using the following formula: ； Based on the array adaptation index Rmatch, the drive plate fine-tuning mechanism completes the setting of the second spacing d2 to achieve individualized adaptation.

6. The intelligent counting and marking method for the insert-type hair follicle transplant device according to claim 5, characterized in that, Step 3: Multi-angle and multi-band image acquisition is performed on both sides of the individually fitted insert to obtain hair follicle structure maps, vascular ring feature maps, and dermal papilla maps. Based on tissue boundary integrity, papilla clarity, and vascular texture features, an activity assessment operation is performed to construct the structural clarity coefficient Cstruct, the hair follicle blood supply feature coefficient Cvasc, and the tissue integrity coefficient Cpt. Specific steps include: S311. The construction of the structural clarity coefficient Cstruct is achieved by converting the hair follicle structure map into a grayscale image, applying edge enhancement processing to the horizontal and vertical directions of the image, extracting the gradient changes of the hair follicle edge regions in the image, and statistically analyzing the average intensity of the edge response in the overall image as the edge intensity E of the structure map. S312. Using the gray-level distribution statistical analysis method, extract the standard deviation of the gray-level distribution based on the amplitude of the gray-level change in the image, and use it as the image contrast C. S313. Use the Canny operator to extract the edge contour of the hair follicle structure in the hair follicle structure image to obtain a set of closed-loop boundary lines. Apply the contour tracking algorithm to extract the main boundary of the hair follicle region. Count the breakpoints and successive segments of the main boundary. Compare the number of breakpoints and successive segments with the total length of the boundary to estimate the degree of boundary damage in the image, which is used as the boundary damage ratio R. S314. After dimensionless processing using the edge intensity E of the structural map, image contrast C, and boundary damage ratio R, the structural sharpness coefficient Cstruct is calculated using the following formula: 。 7. The intelligent counting and marking method for the insert-type hair follicle transplant device according to claim 6, characterized in that, The specific steps for constructing the hair follicle blood supply characteristic coefficient Cvasc are as follows: S321. By standardizing the grayscale of the vascular ring feature map, highlighting the fine vascular structure through the vascular enhancement filter, distinguishing the vascular pixels from the background area using image segmentation technology, counting the number of pixels identified as vascular, and comparing the size of the area containing vascular in the whole image to obtain the vascular density D. S322. Extract the vascular contour by using an edge detection algorithm on the vascular ring feature map, identify all independent vascular structures based on connectivity analysis, count the number of closed-loop structures and the total number of vascular structures, and obtain the vascular closed-loop rate L. S323. By skeletonizing the vascular ring feature map and compressing it into a centerline representation with a single pixel width, all vascular segments are identified, and endpoint pixels in the vascular segment are extracted. The number of vascular breakpoints and the total number of vascular segments are counted to obtain the vascular breakage rate G. S324. Combining vascular density D, vascular closure rate L, and vascular rupture rate G, after dimensionless processing, the hair follicle blood supply characteristic coefficient Cvasc is calculated as follows: 。 8. The intelligent counting and marking method for the insert-type hair follicle transplant device according to claim 6, characterized in that, The specific steps for constructing the organizational integrity coefficient are as follows: S331. By using the Sobel algorithm to perform edge detection on the dermal papilla image, the papilla edges are extracted. Then, through contour tracking and connected component analysis, the edge closure path is identified, and the papilla contour closure degree H is obtained. S332. By performing gray-level co-occurrence matrix analysis on the hair papilla image, local texture features are extracted, the stability of gray-level distribution within the region is analyzed, and blurring, discontinuity, and texture disorder are identified through texture clarity and continuity to obtain internal texture integrity B. S333. The edges of the hair papilla image are extracted using the Sobel algorithm. Then, through morphological processing and contour analysis, edge breakpoints and gaps are identified. The gap area ratio DL is obtained by comparing the gap length with the total contour length. S334. Combining the nipple contour closure H, internal texture integrity B, and notch area ratio DL, after dimensionless processing, the tissue integrity coefficient Cpt is calculated using the following formula: 。 9. The intelligent counting and marking method for the insert-type hair follicle transplant device according to claim 8, characterized in that, Step three also includes: S341. Based on the calculated structural clarity coefficient Cstruct, hair follicle blood supply characteristic coefficient Cvasc, and tissue integrity coefficient Cpt, after dimensionless processing, the hair follicle activity index Factive is calculated using the following formula: ； In the formula, w1, w2, and w3 represent weighting coefficients; S342. By setting a preset hair follicle activity index threshold TF, and comparing the hair follicle activity index Factive with the hair follicle activity index threshold TF, the following evaluation results are obtained: When the hair follicle activity index (Factive) is greater than or equal to the hair follicle activity index threshold (TF), it is marked as "preferred for implantation". When the hair follicle activity index threshold TF When 70% ≤ hair follicle activity index (Factive) < hair follicle activity index threshold (TF), mark it as "routine implantation"; When the hair follicle activity index Factive < the hair follicle activity index threshold TF When the percentage reaches 70%, mark it "Use with caution"; this indicates that the image quality is low or the tissue is severely damaged, and it is recommended to refrain from using it.

10. The intelligent counting and marking method for the insert-type hair follicle transplant device according to claim 9, characterized in that, Step four includes: S41. For hair follicles that may be repeatedly identified in multi-angle images with the same hole HID, perform fusion and normalization processing based on spatial location and image features to ensure that each hair follicle appears only once in the list, forming the final deduplicated hair follicle entry. S42. The deduplicated hair follicle entries are combined with the hair follicle activity index (Factive) markings of "preferred implantation", "routine implantation" and "temporarily postponed use" to automatically generate an intraoperative implantation and allocation list. This list is displayed and filtered in real time through the doctor's operating terminal, assisting the doctor in quickly selecting hair follicles during the hair follicle transplantation stage and calculating the number of hair follicles after the procedure.

Images