KR101644843B1 - Analyses of condition based on the peculiarity of face temperature distribution - Google Patents

Abstract

The present invention relates to a state analysis method using the temperature distribution specificity of a face, and relates to a method of analyzing an individual's personality and health state using a temperature distribution specificity of a face photographed by a thermal camera. The temperature distribution of the face by the thermal camera shows the specific shape of each person by combining with the shape of the face. It accumulates the data in a database and analyzes the change of normal health condition and abnormal health condition through comparison with the database to detect the health condition .

Description

[0001] The present invention relates to a method for analyzing a state of a face using a temperature distribution singularity,

The present invention relates to a method and a device for analyzing a face image of an individual using a thermal camera, in particular, a temperature distribution between a forehead and a lips by using a facial feature, that is, State diagnosis, and a method of state analysis using the temperature distribution specificity of the face, which diagnoses the condition and the health state using the distortion, variation, and defect of the specificity.

All objects whose absolute temperature is not 0 ° generate electromagnetic waves with a wavelength that is related to their temperature, and most objects around us have their center wavelengths in the infrared region.

A thermal camera is a camera that captures images using infrared rays emitted from these objects. Especially in the case of animals, the temperature of the body is controlled by a thermal camera at a resolution of 0.1 ° or less It is possible to shoot.

The temperature distribution on the surface of the human body photographed by such a thermal camera enables the primary diagnosis of all diseases accompanied by high fever. Therefore, it is possible to identify the group of passengers with abnormal signs at the airport, the breast cancer, It is used in many fields such as identification of cause of pain without pain, diagnosis of local health condition like thyroid, nerve damage detection in dentistry, Vein Thrombosis, circulatory disorder of lower limb, , And the application range will be broadened with the recent start of telemedicine.

In these applications, the temperature of the area with inflammation or pain in the human body is higher than the temperature of the normal part, so that the difference in color with the other part is recognized, so that the part is recognized. In a general thermal camera, The high temperature is colorized so that it appears close to saturated white.

This temperature distribution is a random form that does not have a shape and indicates the approximate position of the symptom region. Unusual, however, in many cases, the temperature distribution of the face above the lips, including forehead, is fused with each individual's shape, ie, shape and profile, It is also visible. The shape observed by the temperature profile fused with the facial shape and contour is typically represented as an animal face shape such as a tiger, a raccoon, a fox, a dog, a monkey, a goat, a wolf, a badger, It is confirmed that the shape is different from other specific shapes.

It was also found that even if an animal of the same kind had a different face, the expression of each face varied according to the condition. Therefore, the more samples taken over time, the more likely it is that animal forms with different facial shapes or various facial expressions may be found.

Therefore, if a temperature distribution of a certain region forms a specific shape in a steady state, it is possible to analyze an individual's condition and health state by observing a change of the shape according to an individual's condition or health condition .

Japanese Patent Application Laid-Open No. 2009-119896 (published on June 4, 2009)

It is an object of the present invention to provide a method for diagnosing a health condition of an individual by using a characteristic shape in which a face shape and an outline of a face are combined with a temperature distribution of a face, And to provide a state analysis method using the temperature distribution singularity of a face applicable to a new medical diagnostic method including the present invention.

According to another aspect of the present invention, there is provided a method for analyzing a state using a temperature distribution specificity of a face, the method comprising: capturing a face portion of a subject with a thermal camera to obtain a temperature distribution image; Recognizing an identification region corresponding to a designated face portion of the subject in the temperature distribution image and extracting a temperature distribution for the identification region; Extracting feature information from the temperature distribution image of the entire face of the subject; Receiving status information of a subject; Accumulating the temperature distribution image, the temperature distribution of the identification region, the feature information, and the state information to construct a database; Analyzing a new image obtained by photographing a face portion of a subject with a thermal image camera, and outputting an analysis result reflecting state information with high similarity through searching of the database; .

Receiving a first classification criterion based on an animal shape with respect to the feature information; Comparing the extracted feature information with the first classification reference, calculating a similarity to the first classification reference, and applying a first classification reference having the highest similarity; And each feature information stored in the database may be classified by applying a first classification criterion.

Receiving a second classification criterion based on a facial expression change with respect to the first classification criterion; Comparing the feature information to which the first classification reference is applied to the second classification reference, calculating a similarity to the second classification reference, and applying a second classification reference having the highest similarity; And the feature information stored in the database and to which the first classification criterion is applied may be classified by applying a second classification criterion.

The status information may include identification information of the subject, and identifying the subject through the analysis result. As shown in FIG.

Generating change information by tracking a change in a temperature and a distribution area in a temperature distribution of an identification region for the same subject stored in the database, and storing the change information in the database; Updating the change information as the temperature distribution of the identification region for the same subject is accumulated in the database; .

The characteristics of the human face, which is fused with the temperature distribution, are one of the newly discovered characteristics. The analysis of the condition and the health condition of the subject using these features is based on the fact that the physical and mental Providing a new means of diagnosing a health condition in a simple yet more accurate manner and providing a new medical diagnostic technique that can identify the relationship with health status.

FIG. 1 is a photograph of a thermal camera with pain in the right knee,
FIG. 2 is a photograph of a thermal camera showing an example of a feature formed by fusion of a face shape and an outline with a temperature distribution,
FIG. 3 is a photograph showing a face by a thermal camera of the same person taken at a time difference,
Figure 4 is a photograph of a thermal camera showing a feature that is difficult to associate with a particular animal shape,
FIG. 5 is a photograph of a thermal camera comparing features of two images taken at different times,
6 is a flowchart according to a preferred embodiment of the present invention.

The present invention relates to a method and a device for analyzing a face image of an individual using a thermal camera, in particular, a temperature distribution between a forehead and a lips by using a facial feature, that is, Condition diagnosis, and a new method for diagnosing conditions and health conditions using distortion, variation, and defects of this specificity.

This feature is basically an individual shape that differs from person to person because the face shape, contour and temperature distribution of each person is different. In addition, it may appear as various animal shapes with various facial expressions depending on individuals.

The shape of these shapes is basically unchanged even after a long period of time. However, since these shapes and changes represent various facial expressions according to physical condition and health, they are related to individual identification, individual personality, condition and health condition It is to diagnose the personality, condition and health condition and to judge the health and condition through the detection of the degree of distortion and the damaged part of this shape. Here, "state" means a physical phenomenon in which the degree of relaxation of the individual mental state or the degree of fatigue of the present body is fused individually or simultaneously.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, a method for analyzing states using the temperature distribution specificity of a face of the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a photograph of a thermal camera in a state in which pain is present in the right knee. Generally, an image photographed by a thermal camera shows only a rough part of the affected part as shown in FIG.

1 shows the pain in the right knee. In the right knee (1) and the left knee (2), the difference in image is in the intensity of the heat caused by the pain. Most of the right knee (3).

The degree of saturation of white in the image of FIG. 1 represents the highest temperature as shown in the color diagram (Diagram) (4) given as the right index. This diagram (4) shows that dark blue has the lowest temperature and then the temperature increases in the order of light blue, green, yellow, red and white.

In FIG. 1, since the pain site 3 occupies most of the right knee 1, it is difficult to substantially locate the pain. The heat generated from the epicenter of the actual pain spreads to all directions around the epicenter. However, our human tissue is very uneven, and the epicenter itself has various forms along with the structural characteristics of the surrounding heat spreading. Pain areas with symmetrical patterns are difficult to generate.

FIG. 2 is a photograph of a thermal camera showing an example of a feature formed by fusion of facial shape, contour, and temperature distribution, and is a photograph of nine different facial lacerations obtained from nine individuals.

In the photographs shown in Figs. 2 (a) to 2 (i), circular dotted lines indicate the identification region in the present invention, and include the upper region of the upper lip and the forehead region including the forehead bounded by the hair. This area includes the forehead, eyebrows, eyes, perimeter of the eye, cheekbones, cheekbones, and nose, nostrils, and skin areas on the face. These nine photographs show the wolf shape 5 of Fig. 2 (a), the fox shape 6 of Fig. 2 (b), the raccoon shape 7 of Fig. 2 (c) 2, the monkey shape 9 of Fig. 2 (e), the tiger shape 10 of Fig. 2 (f), the badger shape 11 of Fig. 2 (g) , And another wolf shape 13 of Fig. 2 (h), respectively.

As shown in FIG. 2 (a), the photographs of the photographs are taken at the evening meal (pe = 0.99) at 07:30:07 pm on May 21, 2008, at 07:39 pm on October 02, 58 pm (pe = 0.95), Fig. 2 (c) is the lunch time (e = 0.99) for the lunch on May 21, 2008 at 11:09:34 am, (Pe = 0.99), Fig. 2 (e) shows the evening meal (pe = 0.99) and Fig. 2 (f) (Pe = 0.98), Fig. 2 (g) shows the results of the post-lunch (pe = 0.98) It is 7:13:32 pm evening meal (pe = 0.99).

2 (g) is female, and the photographs of FIG. 2 (a), FIG. 2 (c) and FIG. 2 (h) are obtained from foreigners who are not Koreans. The wolf shape (5) of Fig. 2 (a) and the other wolf shape (13) of Fig. 2 (i) exhibit a similar atmosphere, but their details are different. These forms are characteristics of individuals who are independent of race, and there is no example in which a person has two forms, but it is clear that similar classification is possible, but different depending on the subject.

Of these features, the characteristic feature of FIG. 2 (b) is that there is a white temperature distribution in most of the portions excluding the eyes and nostrils from the forehead to the mouth.

FIG. 3 is a photograph showing a face of a same person taken by a time lag with a thermal camera. As shown in FIG. 3 (a), the photographs of the photographs were taken at 11:55:05 am on May 21, 2008 and peaks of 0.99 (pe: 0.99) (E = 0.99), Fig. 4 (d) is 7: 00 pm on May 21, 2008, 48:15 evening meal (pe = 0.99).

3 (a), which is a thermal image taken at 12:00 am before lunch, shows a monkey-like appearance (14) in a flooded expression, but in Figure 3 (b) 3 (c), which is a thermal image taken at about 5:00 pm, shows a monkey-like appearance 17 with a happy expression on the eyes with eyes 15 and 16 and a size of two eyes 15 and 16 (17), as shown in FIG. 3 (b), because of an increase in the area of the surrounding white area, the tired face of the monkey (18) with eyes that are much smaller than the sizes of the two eyes. In addition, the head part (19) of the monkey figure (18) with a tired expression has a yellow color overall and a relatively high temperature showing a temperature distribution in the red and white parts, but the monkey appearance (17) And the temperature of the head region 20 is basically a light blue color and exhibits a low temperature distribution.

The temperature distribution in the head region 19 of the appearance 18 of the monkey-like tired expression of FIG. 3 (c) is well matched to the expression. The temperature distribution 21 appearing around the nose in the monkey appearance 17 of the happy expression as shown in Fig. 3 (b) is much higher than the temperature distribution 22 appearing around the nose of the monkey appearance 18 of the tired expression But the shape of a large frame, that is, the shape of a particular animal, given by the overall temperature distribution, does not change. This means that the appearance of the feature depends not only on the temperature distribution but also on the partial face appearance of the individual.

The other meaning of FIG. 3 is that the temperature distribution of the face varies somewhat depending on the condition and the health condition, but the basic framework in which the distribution is fused with the feature, that is, Is not changed, but the atmosphere that the specific face expresses by the condition or the health state, that is, the expression is changed. This is not different from the appearance of an individual's face or the atmosphere expressed by his face depending on his or her condition or health condition.

FIG. 4 is a photograph of a thermal camera showing a characteristic feature difficult to relate to a certain animal shape, and shows a feature shape that is difficult to relate instantaneously to a specific animal shape in the face shape of the specific animal of the present invention.

However, the shape (23) shows a bear shape when viewed from a distance, and the shape (24) shows a very strong temperature distribution with a flying eagle shape (25) and it is difficult to distinguish the entire shape, The shape of the fluffy dog (26) is shown.

FIG. 5 is a photograph of a thermal camera comparing two feature images taken at different times. Fig. 5 (a) is a photograph taken on December 12, 2014, Fig. 5 (b) is a photograph taken on May 21, 2008, 6.5 years before the same subject, have. 5 (a) and 5 (b), the nose portion 28 shows a high temperature distribution in the shape 13 of Fig. 5 (b) In the shape 27 of FIG. 3A, the corresponding nose portion 29 has the same temperature as the circumference. This temperature difference is analyzed to be due to the decrease in the temperature of the nose portion 29 due to winter in December.

Fig. 5 (c) is a photograph taken on December 12, 2014, Fig. 5 (d) is a photograph taken on May 08, 2008, 6.5 years before the same subject, . The difference between the two shapes 23 and 30 is that the temperature of the fine portion 31 of the shape 23 of FIG. 5 (d) is lower than the fine portion 32 of the shape 30 of FIG. 5 (c). This difference is analyzed to be given by seasonal differences.

Therefore, FIG. 5 proves that the feature shape shows the same shape with little change over time, and it can be seen that the feature shape represents the unique feature of each individual. Since this feature shows a characteristic that is unique to each individual's face shape, it is possible to identify each individual using this shape.

2 and 5, the high temperature region that determines the feature is the area between the eyes and the nose 33, the forehead 34, the perimeter 35, the under eye 36, the nose 37, The peripheral portion 38, the side portion 39 of the nose, etc., and the temperature distribution 40 of the eyebrows representing the shape of the eyebrows also plays a large role in expressing the feature. Therefore, these sites are feature sites for creating features. Fig. 3 shows that the temperature distribution of these characteristic portions varies depending on the condition.

6 is a flow chart according to a preferred embodiment of the present invention. The present invention basically comprises a thermal camera, a system having data processing means capable of analyzing and storing photographed images by an algorithm, Lt; / RTI >

First, the first step S 110 is a step of capturing a temperature distribution image by capturing a face portion of a subject with a thermal image camera. In the present invention, such a thermal image is continuously collected according to a specific period or a specific situation . The more the collection period and the number of images to be collected, the more advantageous it is, and it is preferable to set the possible ambient temperature and place to similar conditions.

In the next step S 120, an identification region corresponding to a designated face portion of the subject is recognized in the temperature distribution image, and a temperature distribution for the identification region is extracted. At this time, the identification area can be appropriately changed according to needs, but in the preferred embodiment of the present invention, as mentioned above, it is a region between the forehead and the upper lip.

As mentioned above, the high temperature region that determines the feature is the region between the eyes and the nose 33, the forehead 34, the perimeter 35, the under eye 36, the nose 37, (38), the side of the nose (39), and the eyebrow temperature distribution (40), which represents the shape of eyebrows, also play a major role in the representation of features. have.

The next step (S 130) is a step of extracting the feature information appearing from the temperature distribution image of the entire face of the subject, and the feature information includes a feature distribution of the feature distribution Information. Basically, this is done through image analysis such as the position of the boundary of the color appearing on the thermal image and the distance between the boundaries.

In the next step (S 140), the state information of the person who has acquired the temperature distribution image is input. The state information includes information about the subject, such as the degree of relaxation of the subject's disease or health condition and mental state, And the state information such as the degree of fatigue of the input information is more specific, so that the application field of the present invention can be broadened.

In the next step S 150, the temperature distribution image, the temperature distribution of the identification region, the feature information, and the state information are accumulated to construct a database. Such a database is a basic data for the analysis of the thermal images to be provided later, and the more accumulation of information, the more accurate diagnosis can be made.

In the next step S 160, the new image obtained by photographing the face portion of the subject is analyzed by the thermal imaging camera, and the analysis result reflecting the highly similarity state information is output through the search of the database. That is, the state of the current subject can be judged to be the same or close to each other through the state shown in the cumulative data of the database having substantially the same or similarity.

At this time, groups of subjects having a similar temperature profile, that is, skeletons or muscles, having a similar temperature distribution can be classified as having the same constitution and characteristics, so that classification can be performed to obtain more meaningful analysis results. Preferably based on an animal shape that is arbitrarily categorized as mentioned above.

For this purpose, a step S 141 of receiving the first classification reference based on the animal shape with respect to the feature information is performed. Even if a certain animal is not attached, it is possible to group similar characteristics of face shape and temperature distribution as a criterion for classification of a subject having similar characteristics.

Thereafter, in step S 142, the feature information extracted in the previous step S 130 is compared with the first classification reference to calculate the similarity to the first classification reference, and the first classification reference having the highest similarity is applied And each feature information stored in the database is classified according to a first classification criterion.

In addition, as described above, since the facial expression changes according to the state of the subject even if the large shape according to the first classification standard does not change, a step of receiving the second classification standard based on the facial expression change with respect to the first classification standard S 143) is preferably performed. That is, based on the change of facial expression based on the illusion, it can be more easily accessed. In the subsequent step (S 144), the feature information to which the first classification reference is applied is compared with the second classification reference, The similarity is calculated and the second classification reference having the highest similarity is applied. The feature information that is stored in the database and to which the first classification reference is applied is classified by applying the second classification reference.

In addition, the feature in the present invention represents a specific shape peculiar to the subject, and it is possible to obtain accurate personal identification by relatively less influencing factors such as makeup and disguise. Accordingly, the state information includes the identity information of the subject and identifies the subject through the analysis result (S 161), so that the individual can be identified.

In addition, it is necessary to accumulate and manage the temperature change continuously based on the reason of the aging of the body or the like so that more accurate state analysis can be performed by the temperature change trend of the above-mentioned characteristic parts.

To this end, the change information is generated by tracking changes in the temperature and the distribution area in the temperature distribution of the identification region for the same subject stored in the database and stored in the database (S 170) And updating the change information as the temperature distribution accumulates in the database. The detailed state determination can be made using distortion, fluctuation, and defect of the temperature distribution.

It is to be understood that the invention is not limited to the disclosed embodiment, but is capable of many modifications and variations within the scope of the appended claims. It is self-evident.

Claims (5)

In a state analysis method using a face temperature specificity,
A step (S 110) of acquiring a temperature distribution image by photographing a face portion of a subject with a thermal image camera;
Recognizing an identification region corresponding to a portion between a forehead and an upper lip of a designated face of the subject in the temperature distribution image, and extracting a temperature distribution for the identification region (S 120);
(S 130) extracting the feature information according to the shape of the face shape of the subject and the boundary line of the color appearing in the temperature distribution image;
(S 140) receiving status information including a disease and a health status of the subject;
A step (S 141) of receiving a first classification reference based on an animal shape with respect to the feature information;
Comparing the extracted feature information with the first classification reference, calculating a similarity to the first classification reference, and applying a first classification reference having the highest similarity (S 142);
(S 143) receiving a second classification reference based on the facial expression change with respect to the first classification reference;
Comparing the feature information to which the first classification criterion is applied to the second classification criterion to calculate a similarity to the second classification criterion and applying a second classification criterion having the highest similarity (S 144);
(S 150) accumulating the temperature distribution image, the temperature distribution of the identification region, the feature information, and the state information to construct a database;
Analyzing a new image obtained by photographing a face portion of a subject with a thermal image camera, and outputting an analysis result reflecting state information with high similarity through searching the database (S 160); Lt; / RTI >
Wherein each feature information stored in the database is classified by applying a first classification criterion and a second classification criterion.
delete delete The method according to claim 1,
Wherein the status information includes the identity of the subject,
Identifying (S 161) the subject's identity through the analysis result; The method of claim 1, further comprising:
The method according to claim 1,
A step (S 170) of generating change information by tracking changes in temperature and distribution area in a temperature distribution of an identification region for the same subject stored in the database, and storing the change information in the database;
Updating the change information as the temperature distribution of the identification region for the same subject is accumulated in the database (S 180); Wherein the method further comprises the step of analyzing the state of the face using the temperature distribution singularity.

Images