Identifying follicular units
A follicular unit harvesting process receives images for analysis and identifies and highlights clusters of hair. The analysis includes pixel type determination, pixel clustering, cluster classification and cluster highlighting or replacement. The analysis of the pixels is performed concurrently, and the results of the analysis are optionally used for automated treatment planning.
1. A computer-implemented method comprising the steps of:
a) receiving a first image of a body surface having a plurality of hair follicles for harvesting;
b) classifying pixels of the image into hair pixels and non-hair pixels types, based on a partitioning of category values of said hair pixels from category values of said non-hair pixels, the category values being representative of statistical properties of the hair and non-hair pixels;
c) grouping, responsive to the classifying, a collection of proximal classified hair pixels together to form a cluster of hair pixels that is indicative of a follicular unit;
d) categorizing the follicular unit, responsive to the grouping and based on area parameters of pixels related to the cluster of hair pixels, into a type that is indicative of the number of hair follicles in the follicular unit.
2. The method of claim 1 , further comprising:
receiving one or more other images in a live sequence and repeating steps a)-d) for the one or more other images.
3. The method of claim 2 , further comprising:
highlighting the categorized follicular unit by replacing pixels of said categorized follicular unit in the image with pixels indicative of one or more visual cues corresponding to the type that is categorized for the follicular unit.
4. The method of claim 3 , wherein the one or more visual cues are visual cues selected from the group consisting of colors, numbers, shapes and textures.
5. The method of claim 2 , -further comprising:
projecting the live sequence on an augmented reality (AR) glasses or heads up display (HUD).
6. The method of claim 1 , wherein the grouping is based on a flood fill algorithm.
7. The method of claim 1 , wherein the area parameters include an area of the follicular unit, (FUA-d), a maximum area of a 1-hair follicular unit (1-FUA-c) in FUA-d, an axis aligned bounding area (BA-d), and ratios thereof.
8. The method of claim 1 , wherein the classifying step is performed in a parallel manner and is achieved by:
dividing the first image into sub-images, each sub-image comprising at least one pixel and classifying one of the sub images concurrently as at least another of the sub-images is classified, using a graphics processing unit (GPU) in order to increase a speed of classification in comparison to conventional serial classifications.
9. The method of claim 8 , wherein the grouping is performed in a parallel manner using said sub-images and the GPU.
10. The method of claim 1 , further comprising scanning an area of the body surface in order to obtain a 3D point cloud or 3D model of the body surface.
11. The method of claim 2 , calculating a number and the type of follicular units to harvest form the body surface based on any combination of (i) a determined grade of baldness, (ii) a determined number of different types of follicular units on the body surface, (iii) a desired final look and (iv) a defined set of harvesting/implantation rules.
12. The method of claim 1 , further comprising:
detecting a movement of an operator's handpiece on the body surface and moving a robotic arm having a camera to follow the movement in order to automatically capture a live video of the body surface proximal to the operator's handpiece.
13. The method of claim 3 , further comprising displaying a textual report that includes information about the categorized follicular unit and updating the textual report based on the one or more other images.
14. A system comprising at least one processor configured to perform the steps comprising:
a) receiving a first image of a body surface having a plurality of hair follicles for harvesting;
b) classifying pixels of the image into hair pixels and non-hair pixels types, based on a partitioning of category values of said hair pixels from category values of said non-hair pixels, the category values being representative of statistical properties of the hair and non-hair pixels;
c) grouping, responsive to the classifying, a collection of proximal classified hair pixels together to form a cluster of hair pixels that is indicative of a follicular unit;
d) categorizing the follicular unit, responsive to the grouping and based on area parameters of pixels related to the cluster of hair pixels, into a type that is indicative of the number of hair follicles in the follicular unit.
15. The system of claim 14 , wherein the processor is further configured to perform the steps of receiving one or more other images in a live sequence and repeating steps a)-d) for the one or more other images.
16. The system of claim 15 , wherein the processor is further configured to perform the step of highlighting the categorized follicular unit by replacing pixels of said categorized follicular unit in the image with pixels indicative of one or more visual cues corresponding to the type that is categorized for the follicular unit.
17. The system of claim 16 , wherein the one or more visual cues are visual cues selected from the group consisting of colors, numbers, shapes and textures.
18. The system of claim 15 , wherein the processor is further configured to perform the steps of projecting the live sequence on an augmented reality (AR) glasses or heads up display (HUD).
19. A non-transitory computer-readable storage medium storing a program which, when executed by a processor, causes the processor to perform a procedure comprising:
a) receiving a first image of a body surface having a plurality of hair follicles for harvesting;
b) classifying pixels of the image into hair pixels and non-hair pixels types, based on a partitioning of category values of said hair pixels from category values of said non-hair pixels, the category values being representative of statistical properties of the hair and non-hair pixels;
c) grouping, responsive to the classifying, a collection of proximal classified hair pixels together to form a cluster of hair pixels that is indicative of a follicular unit;
d) categorizing the follicular unit, responsive to the grouping and based on area parameters of pixels related to the cluster of hair pixels, into a type that is indicative of the number of hair follicles in the follicular unit.
20. The non-transitory computer-readable storage medium of claim 19 , wherein the processor further performs the steps of receiving one or more other images in a live sequence and repeating steps a)-d) for the one or more other images.