KR101339386B1 - System for gesture recognition using tracking pose of a trainee - Google Patents

Abstract

The present invention relates to a system for gesture recognition using tracking pose of a trainee. According to the present invention, the system for gesture recognition using tracking pose of a trainee includes an image recognition part; a posture analysis part; a control part; and a key operation part. The image recognition part include an IR marker. Therefore, the present invention improves a training effect. [Reference numerals] (104) Key control part;(110) Image recognition part;(120) Posture analysis part;(130) Control part

Description

System for gesture recognition using tracking pose of a trainee}

The present invention relates to a system for recognizing a body motion through a posture tracking of a trainer, and more particularly, to a motion recognition system capable of classifying meaningful gestures from a pattern of motion of a trainer through image recognition.

In general, military, police or related trainees will be given tactical or fire drills necessary for wartime or anti-terrorism, disaster relief, and suppression of protests. You will be able to learn the correct shooting posture and meaningful semaphores (body movements), and this training course is essential for preparation.

On the other hand, when the trainee performs any movements necessary for tactical training or shooting training, it is important to perform the correct movement in the correct posture in many ways, and can not be overlooked that the effective training is achieved in practice.

However, until recently, no method was developed to lead the trainees more efficiently. Therefore, most of the tactical or target training is based only on the naked eye. Since it is not known, the necessity of visually confirming the shape of the training and the training result data is emerging.

The present invention was created to solve the above problems, by analyzing the motion pattern of the trainer participating in the recognition and training of the virtual environment through the image recognition using the IR marker and the IR camera and the meaningful body movement of the trainer from the analysis result ( It is to provide a body movement recognition system through tracking the posture of trainers, which enables to classify gestures and shooting training postures, to manage individual and group training result information, and to promote effective training.

One embodiment of the present invention for achieving the above object, a plurality of groups of IR markers attached to the body of the trainer and an image having a plurality of groups of IR cameras for capturing the image including the IR marker to shoot the movement of the trainer Recognition unit, posture analysis unit for classifying the meaningful body movements of the trainees by analyzing the image signal including the IR markers for the trainees photographed through the image recognition unit, and trainees' individual training result information by receiving the analysis results Body movement recognition system through tracking the position of the trainer including a control unit to accumulate and store.

According to the present invention, it is possible to classify the body motion according to the movement pattern of the trainer through image recognition using the IR camera and the IR marker, and to store and provide the individual training result information as an analysis result, so that each training participant is trained. It is not only able to accurately determine whether the correct and accurate training is performed, but also has the advantage of improving the training cost-effectiveness.

Figure 1 is a block diagram illustrating the overall configuration of the body movement recognition system through the posture tracking of the trainer according to the present invention.
2 is a conceptual schematic diagram of the system of FIG.
3 is a reference diagram illustrating the position of the IR camera installed around the trainer in the system of the present invention.
4A to 4D are reference pictures exemplified for explaining an attachment position of an IR marker used for detecting a trainee's motion in the present invention.
FIG. 5 is a diagram illustrating the types of shooting training postures that are classified by the training pattern of the trainer in the posture analyzing unit.
FIG. 6 is a diagram illustrating the types of handwritten signals classified by the trainer's motion in the posture analyzing unit.
FIG. 7 is a conceptual diagram illustrating the discernment of the position and movement characteristics of the head and hand for classifying the handwritten signal of FIG. 6.

Hereinafter, a configuration and an operation of a gesture recognition system through a posture tracking of a trainer according to an embodiment of the present invention will be described in detail with reference to the accompanying drawings.

It is to be understood that the words or words used in the present specification and claims are not to be construed in a conventional or dictionary sense and that the inventor can properly define the concept of a term in order to describe its invention in the best possible way And should be construed in light of the meanings and concepts consistent with the technical idea of the present invention. Therefore, since the embodiments described in the specification and the configuration shown in the drawings are only one of the most preferred embodiments of the present invention, it is understood that there may be various equivalents and modifications that can replace them at the time of the present application. shall.

1 is a block diagram illustrating the overall configuration of a system for recognizing a body motion through a posture tracking of a trainer according to the present invention, and FIG. 2 is a conceptual schematic diagram of the system of FIG. 1. The image recognition unit 110, a posture analysis unit 120, a control unit 130, and a key operation unit 104 for capturing an image of the image are configured.

The image recognition unit 110 includes a plurality of groups of IR markers attached to the body of the trainer, and a plurality of groups of IR cameras for capturing an image including the IR markers to capture the movement of the trainer. In this case, the IR (Iinfra-Red) camera is preferably an infrared camera, and each IR marker is preferably a spherical marker that reflects infrared rays.

The posture analysis unit 120 classifies meaningful body movements of the trainer by analyzing an image signal including an IR marker for a trainer photographed through the image recognition unit.

The controller 130 receives the analysis result of the posture analyzer and accumulates and stores the training result information for each individual trainer.

The key manipulation unit 104 includes a plurality of keys, and generates and transmits a key manipulation signal for inputting trainer information according to a key input arbitrarily operated by a trainer or an operator to the controller 130.

3 is a reference diagram illustrating the position of an IR camera installed around a trainer in the system of the present invention, for example, assuming that the side where the screen 101 is installed is the front of the training space 11, As illustrated in FIG. 3, the image recognition unit 110 may include first group IR cameras 112a and 112b for capturing the hands and head positions of the trainer and second group IR cameras for capturing the positions of both feet of the trainer. 112c), a third group IR camera 112f for photographing the head position of the trainer.

The first group IR cameras 112a and 112b are respectively installed on the upper left and right sides of the screen 101 to capture an image signal including an IR marker attached to the hand and the head of the trainer.

The second group IR camera 112c captures an image signal including an IR marker installed at the upper center side of the screen and attached to both feet of the trainer.

The third group IR camera 112f is installed in the center of the ceiling rear side of the training space and captures an image signal including an IR marker attached to the head of the trainer.

In particular, the image recognition unit 110 may include the first group IR markers 111a-111c for indicating the position of the head of the trainer, and the second group IR marker for indicating the position of one hand (preferably the left hand) of the trainer. 111d-111g, a third group IR marker 111h for indicating the position of one of the trainers, and a fourth group IR marker 111i for indicating the position of the other of the trainers. 4A to 4D are reference photographs for explaining an attachment position of an IR marker used for detecting a motion of a trainer in the present invention, and the first group IR markers 111a to 111c are illustrated in FIGS. 4A and 4B. As shown, it consists of three IR markers attached to the upper part of the helmet or hat worn by the trainer, and each IR marker is attached to radially equidistant positions about the crown. The second group IR markers 111d-111g consist of four IR markers attached to the circumference of the wrist band worn by the trainer on one wrist as illustrated in FIG. 4C, with each IR marker spaced at equal intervals. Is attached. The third group IR marker 111h and the fourth group IR marker 111i consist of respective IR markers attached forwardly around the respective ankle bands that the trainer wears on both ankles, as illustrated in FIG. 4D.

The posture analyzer 120 analyzes image signals captured by the first group IR cameras 112a and 112b, the second group IR cameras 112c, and the third group IR cameras 112f of the image recognition unit 110. It recognizes IR markers attached to the trainee's body and analyzes the characteristics of the position and movement of each IR marker, the distance between IR markers, and whether or not the IR marker is recognized. Do it. In particular, the posture analysis unit 120 is characterized by the characteristics of the head and hand distance, head and hand position, head and hand angle, hand movement, head and foot distance recognized from the IR markers of each group To analyze, we have preset values for each of the upper / mid / lower of the height, the circle / root of the distance, and the height / lower of the angle, and use the features and the reference to classify the meaningful body movements of the trainer. It becomes possible.

FIG. 5 is a diagram illustrating five types of training postures that are classified from virtual image recognition and a training pattern of a trainer when the posture analysis unit 120 is configured to recognize a virtual video environment. , Knee shooting, first down shooting, second down shooting) and two types of shooting training positions (sitting shooting, incident arc shooting) which are automatically set according to the environmental recognition result of the virtual image. It is a reference drawing which illustrates 9 types of handwritten signals which the posture analysis part 120 classifies from the training pattern of a trainer.

In order to classify five types of shooting training postures (such as standing shot, squat shot, knee shot, first down shot, and second down shot) as illustrated in FIG. IR marker detection to check head height and foot position recognized by IR markers attached to head and foot is checked and it has classification criteria for the following five training positions.

1. If the head is in a high position and both IR markers are detected, sort out the firing posture for "Stand Shoot." When the posture analyzing unit 120 in the present invention is set to recognize the virtual video environment, the shooting training posture for the "standing shooting" is classified when the trainer object image mixed in the virtual environment is placed on the flat surface. It is a shooting training posture, and if a trainer object image is located in a pit of a virtual environment, the shooting training posture for "incident arc shoot" is automatically changed.

2. If the head is in the middle position and both feet 'IR markers are detected, sort out the shooting position for "quat shoot". When the posture analysis unit 120 in the present invention is set to recognize the virtual video environment, the shooting training posture for the "quat shoot" is that the trainer object image mixed in the virtual environment will be located in the wetland or CBR environment. It is classified as a shooting training posture, and if the training object image is located on a flat surface, the training posture for the "sitting and shooting" is automatically changed.

3. If the head is in the middle position and the IR marker on one foot is detected, sort out the shooting position for "knee shoot".

4. If the head is in a low position and both IR markers are detected, sort out the shooting position for "first shoot down."

5. If the head is in a low position and an IR marker on one foot is detected, sort out the shooting training posture for "second shoot down."

The classification criteria for the fire training posture described above are illustrated in detail in Table 1.

Shooting training Stand and shoot Squat Knee shoot Shoot down first 2nd down Foot marker detection both side both side One side both side One side Head position Prize medium medium Ha Ha

FIG. 7 is a conceptual diagram illustrating the position and movement characteristics of the head and hand serving as the classification criteria of the handwritten signal illustrated in FIG. 6, and the posture analyzer 120 according to the present invention has an upper / middle / Lower position, circle / root for head and hand distance, height / lower for head and hand angle, and stop for hand movement, vertical repeat, horizontal repeat, horizontal and vertical repeat, left and right repeat, front and rear repeat Depending on whether or not the trainer's meaningful body movements are classified in detail.

The posture analyzing unit 120 according to the present invention has classification criteria for the following nine handwritten signals.

1. If the head and hand are located in a high and intermediate position, the head and hand are close, and the movement of the hand is analyzed as a feature that stops, "Stop" classifies the motion to the handwritten signal.

2. If the head and hand are located in a high and intermediate position, the head and hand are close, and the movement of the hand is analyzed as a vertical repetition, classify the motion to the handwritten signal "move quickly". Do it.

3. If the head and hands are located in the middle and low positions, the head and hands are far from each other, the head and hand angles are low, and the movements of the hands are repeated from side to side, "move slowly." Classifies gestures to handwritten signals.

4. Handwritten signal "lower posture" when the head and hand are in a low position, the head and hand are far from each other, the head and hand angle is high, and the hand movement is vertically repeated. Classify gestures for.

5. If the head and hands are in high and intermediate positions, and the hand movements are analyzed by horizontal and vertical repetition, "dissipate" classifies gestures to handwritten signals.

6. If the head and hand are in a high position, the head and hand are close, and the movement of the hand is analyzed as a horizontal repeating feature, "note" classifies the body motion to the handwritten signal.

7. If the head and hand are in a high position, the head and hand are far from each other, the head and hand angle is high, and the movement of the hand stops. Classify

8. A handwritten signal of "approach or attack" if the head and hand are in a high position, the head and hand are far from each other, the head and hand angle is low, and the hand movement is repeated before and after. Classify gestures for.

9. The head and hand position is high, the head and hand distance is far, the head and hand angle is low, the movement of the hand is stopped, and the impact position coordinates by the gun trigger are analyzed. In this case, handwritten signal for "fire direction" is classified.

The classification criteria for the eight handwritten signals described above are illustrated in detail in Table 2.

Handwriting Head and hand position Street of head and hands Head and hand angle Hand movement Stop Season nearness stop Move quickly Season nearness Vertical repeat Move slowly Medium mum lowness Left and Right Repeat Lower your posture Ha mum height Vertical repeat Dissipate Season Horizontal Vertical Repeat Pay attention Prize nearness Horizontal repeat Keep an eye on Prize mum height stop Approach or target Prize mum height Left and Right Repeat

The controller 130 receives the trainer information from the key input signal that the trainer or the operator arbitrarily manipulates through the key manipulation unit 104, and authenticates the trainer information.

In addition, the controller 130 receives a result of a meaningful motion analysis of the trainees classified by the image recognition in the posture analyzer 120 and receives the result and accumulates the result as individual training result information.

As described above, the system of the present invention has been described with reference to a specific embodiment, but the present invention is not limited to the above embodiment, but is a recording medium including instructions executable by a computer, such as a program module executed by a computer. It may also be implemented in the form of. In addition, it is possible for those skilled in the art to various modifications and variations from this description. Accordingly, it is intended that the scope of the invention be defined by the claims appended hereto, and that all equivalent or equivalent variations thereof fall within the scope of the present invention.

101: screen 104: key control unit
110: image recognition unit 111a-111i: IR marker
112a-112c, 112f: IR camera 120: posture analysis unit
130:

Claims (23)

An image recognizing unit including a plurality of groups of IR markers attached to the body of the trainer and a plurality of groups of IR cameras for capturing an image including the IR markers to capture the movement of the trainer;
A posture analyzer configured to classify meaningful motions of the trainer by analyzing an image signal including an IR marker for the trainer photographed through the image recognizer;
A control unit which receives the analysis result of the posture analyzing unit and accumulates and stores the training result information for each individual trainer; And
It is configured to include; key operation unit for generating a key operation signal for the trainee personal information input to the control unit;
Wherein the image recognizing unit comprises:
The first group IR marker for indicating the position of the trainer's head, the second group IR marker for indicating the position of the trainer's one hand, the third group IR marker for indicating the position of the trainer's one foot and the other foot of the trainer's A fourth group IR marker for indicating location;
The posture analysis unit,
Analyze the position and movement of the IR markers of each group, the distance between the IR markers, and whether the IR markers are recognized or not, and indicate the head height and the position of the feet recognized by the IR markers attached to the head and foot. It checks the presence / absence of IR marker detection to classify meaningful body movements of the trainer,
Upper / middle / lower position relative to head and hand position;
Distant and close to the distance of the head and hands;
Pitch for head and hand angles; And
The body movement recognition system according to the posture tracking of the trainer, characterized in that the classification of meaningful movements of the trainer according to whether the movement of the hand motion, the vertical repetition, horizontal repetition, horizontal and vertical repetition, left and right repetition, front and rear repeat.
delete delete delete delete delete The method of claim 1,
Wherein the image recognizing unit comprises:
A first group IR camera installed at both upper left and right sides of the screen to capture an image signal including an IR marker attached to a trainer's hand and head;
A second group IR camera installed at an upper center side of the screen to capture an image signal including an IR marker attached to both feet of the trainer;
A third group IR camera installed in the center of the ceiling rear side of the training space and capturing an image signal including an IR marker attached to the head of the trainer; .
delete delete delete delete delete delete delete delete delete delete delete delete delete delete delete delete

Images