Method for identifying body constitution in traditional chinese medicine, and computing and processing device and non-transitory computer-readable medium thereof
The present disclosure provides a method and an apparatus for identifying a body constitution, an electronic device, a storage medium and a program, which relates to the technical field of computers. The method includes: acquiring an infrared human-body image of a target user; inputting the infrared human-body image into an infrared critical-point identifying model, to obtain a Traditional-Chinese-Medicine human-body critical point in the infrared human-body image; according to a temperature-distribution profile in the infrared human-body image, determining a temperature type of the Traditional-Chinese-Medicine human-body critical point in the infrared human-body image; and according to the temperature type, identifying a Traditional-Chinese-Medicine body constitution of the target user.
1. A method for identifying a body constitution in Traditional Chinese Medicine, wherein the method comprises:
acquiring an infrared human-body image of a target user;
inputting the infrared human-body image into an infrared critical-point identifying model, to obtain a Traditional-Chinese-Medicine human-body critical point in the infrared human-body image;
according to a temperature-distribution profile in the infrared human-body image, determining a temperature type of the Traditional-Chinese-Medicine human-body critical point in the infrared human-body image; and
according to the temperature type, identifying a Traditional-Chinese-Medicine body constitution of the target user,
wherein the temperature type comprises at least one of a high-temperature types and a low-temperature type; and
the step of, according to the temperature-distribution profile in the infrared human-body image, determining the temperature type of the Traditional-Chinese-Medicine human-body critical point in the infrared human-body image comprises:
determining maximum-temperature-value positions of a first quantity and/or minimum-temperature-value positions of a second quantity in the infrared human-body image; and
determining the temperature type of the Traditional-Chinese-Medicine human-body critical point that corresponds to an image region where the maximum-temperature-value positions are located as the high-temperature type, and/or determining the temperature type of the Traditional-Chinese-Medicine human-body critical point that corresponds to an image region where the minimum-temperature-value positions are located as the low-temperature type; and
wherein the step of determining the maximum-temperature-value positions of the first quantity and/or the minimum-temperature-value positions of the second quantity in the infrared human-body image comprises:
by using a maximum-value filter, acquiring image positions where maximum temperature values of the first quantity are located in the infrared human-body image as the maximum-temperature-value positions;
by using a mean value between a maximum temperature value and a minimum temperature value in the infrared human-body image as an inversion plane, inverting temperature values in the infrared human-body image; and
by using the maximum-value filter, acquiring positions where maximum temperature values of the second quantity are located in the infrared human-body image that has been inverted as the minimum-temperature-value positions.
2. The method according to claim 1 , wherein the infrared critical-point identifying model is obtained by the following steps:
acquiring a natural-light critical-point identifying model and a sample infrared human-body image;
marking the Traditional-Chinese-Medicine human-body critical point in the sample infrared human-body image; and
performing transfer learning of the natural-light critical-point identifying model by using the sample infrared human-body image that has been marked, to obtain the infrared critical-point identifying model.
3. The method according to claim 2 , wherein the step of performing transfer learning of the natural-light critical-point identifying model by using the sample infrared human-body image that has been marked, to obtain the infrared critical-point identifying model comprises:
based on the Traditional-Chinese-Medicine human-body critical point, adjusting fully-connected-layer nodes of the natural-light critical-point identifying model; and
while a model parameter of a feature-extraction layer is maintained unchanged, by using the sample infrared human-body image that has been marked, training the natural-light critical-point identifying model whose nodes have been adjusted, to obtain the infrared critical-point identifying model.
4. The method according to claim 3 , wherein original human-body critical points of the natural-light critical-point identifying model comprise the Traditional-Chinese-Medicine human-body critical point, and a quantity of the original human-body critical points is greater than a quantity of the Traditional-Chinese-Medicine human-body critical point; and
the step of, based on the Traditional-Chinese-Medicine human-body critical point, adjusting the fully-connected-layer nodes of the natural-light critical-point identifying model comprises:
among the fully-connected-layer nodes of the natural-light critical-point identifying model, deleting a fully-connected-layer node whose node label corresponds to a human-body critical point that is not the Traditional-Chinese-Medicine human-body critical point.
5. The method according to claim 3 , wherein the step of, according to the temperature type, identifying the Traditional-Chinese-Medicine body constitution of the target user comprises:
in a Traditional-Chinese-Medicine-body-constitution mapping relation, inquiring Traditional-Chinese-Medicine body constitutions that match with temperature types of the Traditional-Chinese-Medicine human-body critical points.
6. The method according to claim 4 , wherein the step of, according to the temperature type, identifying the Traditional-Chinese-Medicine body constitution of the target user comprises:
in a Traditional-Chinese-Medicine-body-constitution mapping relation, inquiring Traditional-Chinese-Medicine body constitutions that match with temperature types of the Traditional-Chinese-Medicine human-body critical points.
7. The method according to claim 2 , wherein the step of performing transfer learning of the natural-light critical-point identifying model by using the sample infrared human-body image that has been marked, to obtain the infrared critical-point identifying model comprises:
based on the Traditional-Chinese-Medicine human-body critical point, adjusting fully-connected-layer nodes of the natural-light critical-point identifying model; and
by using the sample infrared human-body image that has been marked, training the natural-light critical-point identifying model whose nodes have been adjusted, to obtain the infrared critical-point identifying model.
8. The method according to claim 7 , wherein the step of, according to the temperature type, identifying the Traditional-Chinese-Medicine body constitution of the target user comprises:
in a Traditional-Chinese-Medicine-body-constitution mapping relation, inquiring Traditional-Chinese-Medicine body constitutions that match with temperature types of the Traditional-Chinese-Medicine human-body critical points.
9. The method according to claim 2 , wherein a feature-extraction network in the natural-light critical-point identifying model is a lightweight feature-extraction network, wherein a parameter data volume of the lightweight feature-extraction network is less than a parameter data volume of an original feature-extraction network of the natural-light critical-point identifying model.
10. The method according to claim 9 , wherein the step of, according to the temperature type, identifying the Traditional-Chinese-Medicine body constitution of the target user comprises:
in a Traditional-Chinese-Medicine-body-constitution mapping relation, inquiring Traditional-Chinese-Medicine body constitutions that match with temperature types of the Traditional-Chinese-Medicine human-body critical points.
11. The method according to claim 2 , wherein the Traditional-Chinese-Medicine human-body critical point comprises at least one of a head critical point, a neck critical point, a shoulder critical point, an elbow critical point, a hand critical point, an abdomen critical point, a crotch critical point, a knee critical point and a foot critical point.
12. The method according to claim 11 , wherein the step of, according to the temperature type, identifying the Traditional-Chinese-Medicine body constitution of the target user comprises:
in a Traditional-Chinese-Medicine-body-constitution mapping relation, inquiring Traditional-Chinese-Medicine body constitutions that match with temperature types of the Traditional-Chinese-Medicine human-body critical points.
13. The method according to claim 2 , wherein the step of, according to the temperature type, identifying the Traditional-Chinese-Medicine body constitution of the target user comprises:
in a Traditional-Chinese-Medicine-body-constitution mapping relation, inquiring Traditional-Chinese-Medicine body constitutions that match with temperature types of the Traditional-Chinese-Medicine human-body critical points.
14. The method according to claim 1 , wherein the step of, according to the temperature type, identifying the Traditional-Chinese-Medicine body constitution of the target user comprises:
in a Traditional-Chinese-Medicine-body-constitution mapping relation, inquiring Traditional-Chinese-Medicine body constitutions that match with temperature types of the Traditional-Chinese-Medicine human-body critical points.
15. A computing and processing device, wherein the computing and processing device comprises:
a memory storing a computer-readable code; and
one or more processors, wherein when the computer-readable code is executed by the one or more processors, the computing and processing device executes the method for identifying a body constitution in Traditional Chinese Medicine according to claim 1 .
16. A non-transitory computer-readable medium, wherein the computer-readable medium stores a computer program of the method for identifying a body constitution in Traditional Chinese Medicine according to claim 1 .