KR101235525B1 - Homing navigation method of mobile robot based on vision information - Google Patents

Abstract

The present invention relates to a homing navigation method of a traveling robot using visual information, comprising: a target map generation process of generating a target map using landmark information acquired by the traveling robot performing unit movement at a target point; And a homing direction determination process of determining a direction for the traveling robot to return to the target point at an arbitrary position.

Description

Homing navigation method of driving robot using visual information {HOMING NAVIGATION METHOD OF MOBILE ROBOT BASED ON VISION INFORMATION}

The present invention relates to a homing navigation method of a traveling robot using visual information. More specifically, the distance from the traveling robot to a landmark can be estimated using only visual information, and the traveling robot is provided with a direction sensor at an arbitrary position. Even if not, the present invention relates to a technique for determining a return direction by accurately identifying a location using an arrangement of landmarks.

Many insects and animals find their way and go home in their own way. They use a variety of senses such as sight, smell, touch, and hearing. The most popular sensory information is visual information, which includes insects such as desert ants, bees, and mammals such as mice. It is known to look.

It is not known exactly what kind of algorithms insects actually use to focus on what kind of visual information, but various methods have been proposed and studied. In particular, an image matching method using a snapshot model has been widely studied.

However, the conventional method using the visual information has the disadvantage of requiring a complicated calculation, or a separate device such as GPS to determine the direction.

Therefore, the present invention is to solve the above problems of the prior art, it is possible to estimate the distance from the traveling robot to the landmark using only the visual information, even if the traveling robot at any position does not have a direction sensor It is an object of the present invention to provide a homing navigation method for a traveling robot using visual information to accurately determine a location by using an array of landmarks to determine a homing direction.

In the homing navigation method of a traveling robot using visual information according to an embodiment of the present invention for achieving the above object, a target map is generated using landmark information acquired by a traveling robot performing unit movement at a target point. Goal map generation process; And a homing direction determination process of determining a direction for the traveling robot to return to the target point at an arbitrary position.

Here, the unit motion of the traveling robot means to take a snapshot of the surrounding environment before and after the traveling robot moves by a unit distance.

The driving robot estimates the distance to the landmark by using the landmark information taken on the two snapshots taken through the unit movement, and in particular, the driving robot measures the landmark movement amount between the two snapshots. The distance to the landmark can be estimated.

In addition, the target map preferably includes distance and angle information from the target point to the landmark.

The homing direction determining process may include: obtaining, by the traveling robot, the landmark information by performing unit movement at the arbitrary position; And estimating the position of the traveling robot by projecting landmark information acquired at the arbitrary position on the target map.

Here, in the estimating the position of the driving robot, in projecting the landmark information obtained at the arbitrary position on the target map, the projection order is rotated by the number of the landmarks and the landmark on the target map is rotated. It is desirable to determine the degree of convergence of the projected landmark vectors by matching, and select the arrangement that most similarly converges to determine the direction to the target point.

In the estimating the position of the robot, it is preferable to determine the degree of convergence except for the one with the largest dispersion among the landmark vectors.

Meanwhile, the distance to the estimated landmark may be discretized in a predetermined step and used to determine the return direction.

According to the present invention, it is possible to implement navigation using only one visual sensor without complicated calculations or additional equipment, and thus it is possible to implement the driving robot cheaper and more efficiently.

In addition, according to the present invention, it is possible to store only the information at the target point without having to store a large amount of visual information and compare the present information with the current visual information to obtain a desired result, which is a great advantage in terms of memory.

In other words, the present invention is economical compared to the conventional navigation technology in terms of sensors, memory, and information processing, and when applied to a traveling robot, home navigation can be realized at a lower cost.

1 is a flowchart of a homing navigation method of a traveling robot using visual information according to the present invention;
2 is a conceptual diagram illustrating a process of obtaining landmark information by a traveling robot performing unit movement;
3 is a diagram illustrating an example of a target map generated by the present invention, and
4 is a conceptual diagram illustrating a process of determining a return direction through rotation of a landmark projection order.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art may easily implement the present invention. In the following detailed description of the preferred embodiments of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted when it may make the subject matter of the present invention rather unclear. In the drawings, like reference numerals are used throughout the drawings.

In addition, in the entire specification, when a part is referred to as being 'connected' to another part, it may be referred to as 'indirectly connected' not only with 'directly connected' . Also, to "include" an element means that it may include other elements, rather than excluding other elements, unless specifically stated otherwise.

The present invention is based on a navigation method through image matching that appears in insects such as desert ants and bees, and in particular, by introducing a distance concept to an image matching method using landmarks, a more effective and highly applicable homing navigation method is proposed. .

According to the present invention, the traveling robot first obtains and stores the distance and angle from the target point to the landmark, and then, if the user wants to return to the target point from any position, the distance from the arbitrary position to the landmark in the same manner. Obtain and save the angle. The traveling robot determines the direction to the target point by projecting the information acquired at an arbitrary position onto the information obtained at the target point.

To this end, the distance from the traveling robot to the landmark must be estimated using only visual information, and even if the direction sensor is not provided, the position can be estimated using the arrangement of the landmarks.

In the present invention, the traveling robot performs unit movement to estimate the distance from the traveling robot to the landmark using only visual information. Here, the unit motion of the traveling robot means that the traveling robot takes a snapshot of the surrounding environment at an arbitrary position, moves by a unit distance, and then takes a snapshot of the surrounding environment again. The traveling robot estimates the distance from the traveling robot to the landmark by using the landmark information captured in two snapshots acquired through unit movement.

In addition, the traveling robot uses an array of landmarks to estimate the position and determine the direction to the target point without using a direction sensor. Specifically, in the process of projecting the information acquired at an arbitrary position to the information obtained at the target point, the traveling robot selects an arrangement in which the estimated positions converge most similarly by rotating the landmark projection order, and then directs the direction to the target point. Will be decided.

Hereinafter, the homing navigation method of the traveling robot according to the present invention will be described in more detail with reference to FIGS. 1 to 4.

1 is a flowchart of a homing navigation method of a traveling robot using visual information according to the present invention.

As shown in FIG. 1, the homing navigation method according to the present invention includes a target map generation process (S10) and a homing direction determination process (S20), both of which are configured to infinitely repeat the unit motion of the traveling robot. do.

Here, the unit motion of the traveling robot refers to taking a snapshot of the surrounding environment before and after the traveling robot moves by a unit distance as described above, and the traveling robot uses the information of the landmarks captured in the two snapshots. Estimate the distance to the landmark. Specifically, the traveling robot may measure the distance from the traveling robot to the landmark by measuring the amount of landmark movement between two snapshots.

First, in the target map generation process (S10), the driving robot performs unit movement at the target point to obtain landmark information and generates a target map using the acquired landmark information. In this case, the target map includes distance and angle information to the landmark.

2 is a conceptual diagram illustrating a process of obtaining landmark information by a traveling robot performing unit movement.

이고, T에서 랜드마크까지의 각도는

임을 알 수 있다. 이 경우, 주행 로봇이 단위 움직임 시 이동하는 단위 거리 d를 알면, C 및 T에서 랜드마크까지의 각도를 사인 법칙에 적용하여 수학식 1에 따라 랜드마크까지의 거리를 추정할 수 있다.
Referring to FIG. 2, C represents a position before the traveling robot performs unit movement, and T represents a position after the traveling robot performs unit movement. And from the snapshot, the angle from C to the landmark is

Where the angle from T to the landmark is

. In this case, if the traveling robot knows the unit distance d moving during the unit movement, the distance to the landmark can be estimated by applying the angles from C and T to the landmark to the sine law.

According to the above method, the traveling robot can obtain the distance to the landmark without configuring stereo vision using a distance sensor, an infrared sensor, or an additional camera. In other words, the driving robot may acquire landmark information using only one omnidirectional camera and generate a target map using the same.

3 is a diagram illustrating an example of a target map generated by the present invention.

Referring to FIG. 3, a red landmark represents a location of a landmark stored in a target map. In addition, the landmark vectors illustrated in FIG. 3 represent vectors created when the traveling robot views the landmarks at an arbitrary position.

On the other hand, the homing direction determination process (S20) is a process of determining the direction for returning to the target point from any position after the running robot has completed the environmental search.

To this end, the traveling robot obtains landmark information by performing unit movement at an arbitrary location, and projects the acquired landmark information on a target map to estimate its location and determine a return direction. At this time, the traveling robot can accurately grasp the direction to the target point without any equipment through the landmark projection sequence rotation.

The method of obtaining landmark information by performing the unit movement at an arbitrary position by the traveling robot is the same as described above in the target map generation process S10, and a detailed description thereof will be omitted.

Also, in order to estimate the position of the landmark using only visual information without using a direction sensor, the traveling robot may perform the projection order as many as the number of landmarks in projecting the landmark information acquired at an arbitrary position on the target map. By rotating and matching the landmarks on the target map to determine the degree of convergence of the projected landmark vectors, the most similarly convergent array is selected to determine the direction to the target point. In addition, it is desirable to calculate the convergence point except for the one with the largest variance among the landmark vectors, in order to reduce the influence of noise and errors resulting from the incorrectly extracted landmark.

4 is a conceptual diagram illustrating a process of determining a return direction through rotation of a landmark projection order.

In the case of (a) of FIG. 4, it can be seen that the position estimation using the landmark vector converges most similarly to the landmark vector shown in FIG. 3 as compared with (b) and (c). As such, the position estimation through the rotation of the landmark projection order enables the estimation of the return direction by accurately estimating the position even when there is no direction sensor. In addition, even if there is noise in the landmark information of the snapshot or there is an error in the extraction of the landmark, the effect can be greatly reduced by rotating the landmark projection order.

In addition, when estimating the distance from the traveling robot to the landmark by performing unit motion, it is also possible to discretize the estimated distance. In fact, the insects can be used by discretizing the estimated distance in several predetermined steps in the present invention in consideration of determining that the relative distance is close and close with their own criteria rather than accurately measuring the distance to the landmark. For example, in the case of discretizing the estimated distance in three steps, the distance from the running robot to the landmark may be divided into 'far / middle / near' and used as distance information.

The present invention is not limited to the above-described embodiments and the accompanying drawings. It will be apparent to those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.

S10: Goal Map Generation Process
S20: Return Direction Determination Process

Claims (9)

delete delete delete delete delete delete A target map generation process of generating a target map by using the landmark information acquired by the driving robot performing unit movement at a target point; And
A homing direction determining process of determining a direction for the traveling robot to return to the target point at an arbitrary position;
The homing direction determination process,
Obtaining, by the traveling robot, the landmark information by performing unit movement at the arbitrary position; And
Estimating the position of the traveling robot by projecting landmark information acquired at the arbitrary position on the target map;
Estimating the position of the traveling robot,
In projecting the landmark information acquired at the arbitrary position on the target map, the projection order is rotated by the number of landmarks and matched to the landmark on the target map to determine the degree of convergence of the projected landmark vectors. And selecting an array that most similarly converges to determine a direction to the target point.
The method of claim 7, wherein
Estimating the position of the traveling robot,
A homing navigation method for a traveling robot, characterized in that the convergence degree is determined except for the one with the largest dispersion among the landmark vectors.
A target map generation process of generating a target map by using the landmark information acquired by the driving robot performing unit movement at a target point; And
A homing direction determining process of determining a direction for the traveling robot to return to the target point at an arbitrary position;
The unit motion of the traveling robot means to take a snapshot of the surrounding environment before and after the traveling robot moves by a unit distance, and the driving robot takes landmark information recorded on two snapshots taken through the unit movement. To estimate the distance to the landmark,
Homing navigation method according to claim 1, wherein the distance to the estimated landmark is discretized in a predetermined step and used to determine the homing direction.

Images