TWI384407B - Region identification system by using braille code design - Google Patents

Description

Area recognition system using Braille coding patterns

The present invention relates to an area recognition system, and more particularly to an area recognition system for positioning using a Braille coding pattern.

With the gradual integration of home robots into human daily life, the demand for home robots is increasing, especially autonomous robots with partial artificial intelligence are an important direction for future research and development. For example, these home or autonomous robots can be set to sweep their own homes and handle household chores, or set up simulations for pets to accompany their families, and even robots can be used to carry goods in the factory. Regardless of the function of the robot, the first thing to do is to determine where it is, and then perform various functions depending on the program settings. Therefore, the robot-based area recognition system has been one of the key directions for research and development improvement for many years.

Table 1 lists the positioning techniques and constraints used by the three robot-based area recognition systems, which are briefly described below. The number 1 robot will be equipped with a laser ranging device. By emitting the reflected laser back by a barrier (such as a wall), the distance between the robot and the obstacle can be calculated to locate its position. However, the configuration cost of the laser ranging device is quite high, making the robot positioning system difficult to popularize. Furthermore, when the barrier is a high transmittance material such as glass or a high reflectivity material such as a mirror, the laser ranging device cannot effectively sense the glass or the mirror. As a result, the robot may collide with the glass or the mirror to cause damage.

The robot area identification system of No. 2 uses Radio Frequency Identification (RFID) technology for positioning, wherein the conventional technology configures a reader (Reader) on the robot and configures a plurality of electronic tags (Tag). At a specific location, the inside of the electronic tag stores information about the location. When the robot is close to an electronic tag, the reader reads the information of the location of the tag, thereby positioning the relative position of the robot. However, the configuration cost of the reader and the electronic tag is not cheap, and the robot area recognition system is not easily popularized. Moreover, the communication distance between the reader and the electronic tag is not long, but generally only a few centimeters to several meters. Therefore, when the robot is in a very large space, it is necessary to increase the density of the electronic label, and thus the construction cost of the area identification system needs to be increased.

In addition, the common shortcoming of the two regional identification systems described above is that the price is too expensive. However, in order to reduce the installation cost, the prior art also proposes to directly analyze the environmental image by software. Take the robot numbered 3 as an example, which directly configures a Charge Coupled Device (CCD) to detect external images. The software determines the relative position of the robot by judging the specific target in the image.

However, software identification can only distinguish specific targets such as windows, doors or beds, tables, etc., the main purpose of which is to avoid obstacles or pass the door to another interval, but still can not effectively recognize that the robot is in position. That room (especially when the door or window of the room is the same style). When the target in the image is not much different from the background, the recognition software cannot distinguish the target in the image, resulting in the inability to locate. In addition, when the charge coupling element picks up the bed and the image angle of the table is different, the image deformation is also caused and the software is misjudged.

In addition, when the environment around the robot is too monotonous (such as a repeating pattern of wallpaper), it will cause confusion in the software identification, and thus can not effectively locate. That is to say, there is still too much difficulty in directly performing geometric analysis of surrounding environmental images by software, and the error recognition rate is too high.

In view of the above, the object of the present invention is to provide a region identification system, which is capable of greatly improving the image recognition success rate of a software body by using a simple Braille encoding with a highly reflective material and a reflective light source, and is not susceptible to image deformation angle deformation. problem.

To achieve the above or other objects, the present invention provides an area identification system including a Braille coding pattern and a mobile device, and the mobile device further includes a body, a central control unit, a light source generator, and an image receiver, wherein the central control unit is adapted to control The body moves, and the light source generator and the image receiver are disposed on the body. The Braille coding pattern has area identification information, and the material of the Braille coding pattern is a highly reflective material. The Braille encoding pattern is adapted to reflect the light source to the image receiver when the light source generator produces a light source that illuminates the Braille encoding pattern. In addition, the image receiver is adapted to receive the image of the Braille encoded pattern and transmit it to the central control unit such that the central control unit locates the position of the body based on the area identification information.

In an embodiment of the present invention, the Braille encoding of the Braille encoding pattern may be Chinese encoding, English encoding, Japanese encoding, or Korean encoding.

In an embodiment of the present invention, the Braille encoding pattern may include a frame mark and a Braille code mark, wherein the Braille code mark is disposed in the frame mark. In addition, the Braille coded indicia can be circular, square, diamond, triangular or other suitable shape.

In an embodiment of the invention, the Braille encoding pattern is for reflecting visible light or invisible light, and the high reflective material may be reflective particles, reflective pigments, reflective paint or other suitable materials. In addition, the mobile device further includes a light source generator, and the light source generator is disposed on the body, wherein the light source generator can be a projection lamp, a light emitting diode, a halogen lamp, a fluorescent lamp or a spotlight.

In an embodiment of the invention, the image receiver may be a charge coupled device, a Complementary Metal-Oxide-Semiconductor (CMOS), a spherical camera (PTZ), or a digital camera (DVR).

In summary, in the area recognition system of the present invention, when the image receiver receives an area with high brightness in the image, the recognition software starts to determine whether the high-brightness areas are Braille coding patterns. Since the Braille coding pattern has a simple geometric composition, the central control unit can easily recognize whether there is a Braille coding pattern in the image, and then obtain corresponding area identification information, thereby effectively improving the software identification. Success rate.

The above and other objects, features and advantages of the present invention will become more <RTIgt;

1A is a block diagram of a region recognition system based on a mobile device in accordance with an embodiment of the present invention, and FIG. 1B is a perspective view of the mobile device of FIG. 1A, and FIG. 1C is a front view of the Braille encoding pattern of FIG. 1A. Referring to FIGS. 1A-1C, the area identification system 100 of the present invention includes a mobile device 110 and a Braille encoding pattern 120, wherein the Braille encoding pattern 120 is area identification information that is disposed in a specific area and has this area. In the indoor environment, the Braille code pattern 120 can be attached to the living room, the corridor, the wall of the bedroom, or can be painted on the surface of an obvious item such as a locker, wherein the Braille code pattern 120 in different areas has different Area identification information.

In the above, the mobile device 110 can be a robot, and the mobile device 110 includes a body 112, a central control unit 114, an image receiver 116, and a light source generator 118, and the central control unit 114 is used to control the actuation of the body 112, and the image is received. The device 116 and the light source generator 118 are disposed on the body 112. The light source generator 118 is for providing the light source 118a and illuminating the light source 118a to the outside, and the image receiver 116 continuously captures the external image to transmit the external image to the central control unit 114. The central control unit 114 analyzes the external image by the analysis software, determines whether the external image content includes the Braille encoding pattern 120, and then analyzes the image of the Braille encoding pattern 120 to parse the region identification information. In this way, the central control unit 114 can know the area where the body 112 of the mobile device 110 is located, and can locate the body 112.

The Braille coding pattern 120 is made, for example, of a highly reflective material such as a reflective particle, a reflective pigment, or a reflective paint. In the present embodiment, the Braille coding pattern 120 highly reflects the light source 118a of a specific range of angles. After the image receiver 116 captures the image, the software identification function of the central control unit 114 analyzes whether the image contains a high brightness (this high brightness refers to the brightness of the background). If the image contains high-brightness pixels, the software identification analyzes whether the high-brightness pixels constitute the Braille encoding pattern 120, and further analyzes the area identification information. In this way, the present invention utilizes the highly reflective Braille encoding pattern 120 to generate a high-brightness pixel for the received image, thereby greatly reducing false positives. And effectively improve the correct rate of identification.

FIG. 1D is a diagram showing a plurality of auxiliary lines for further clarifying the Braille encoding pattern of FIG. 1C, and FIG. 2 is an English-encoded Braille encoding table. Referring to FIGS. 1C-1D and FIG. 2, the Braille encoding pattern 120 of the present embodiment includes a frame mark 122 and a Braille encoding mark 124, wherein the Braille encoding mark 124 is located at a specific position within the frame mark 122 to indicate specific information. .

In the present embodiment, each Braille encoding pattern 120 can contain the amount of information of two English letters. In detail, the area within the frame mark 122 can be divided into a left half and a right half, and each half represents an English letter. In addition, each half can be further divided into 2 x 3 total of six bits, and each bit can be Braille encoded mark 124 or blank. In this way, referring to the Braille encoding table of FIG. 2, it is known that the Braille encoding pattern 120 is represented as mb, and the embodiment can define the region identification information represented by mb as the "main bedroom". Therefore, when the central control unit 114 parses the area identification information, the body 112 of the mobile device 110 can be positioned in the "main bedroom".

In view of the above, the Braille coding pattern 120 is very inexpensive to manufacture. In this embodiment, the Braille coding pattern 120 with different area identification information is directly cut by the reflective cloth, and the Braille coding pattern 120 is attached to the wall, the ceiling, the floor or the item, so that the body of the mobile device 110 can be 112 can locate its position no matter where it is.

Because Braille encoding has the advantages of simple structure and the structural features of Braille encoding is very obvious, the success rate and correctness of identification can be greatly improved when geometric image analysis is performed by software design. In addition, the Braille encoding pattern 120 can maintain its structural integrity regardless of the angle at which the image receiver 116 captures the image to cause the Braille encoding pattern 120 to be deformed.

FIG. 3A is a schematic diagram of the image captured by the Braille encoding pattern of FIG. 1C at different angles, and FIG. 3B is a diagram showing a plurality of auxiliary lines for further explaining the resolution process of the Braille encoding pattern of FIG. 3A. Referring to FIGS. 3A and 3B, the image represented by the Braille encoding pattern 120a of FIG. 3A is nearly trapezoidal due to the relationship of the captured image angles. The central control unit 114 first fills the outer edge of the frame-shaped mark 122a of the Braille encoding pattern 120a into a rectangle, and obtains the length L and the width W of the rectangle. In addition, the central control unit 114 may also calculate the pixel area A occupied by the frame-shaped mark 122a, and divide the pixel area A by the length L and then divide by the width W to obtain the first determination value C1:

When the first judgment value C1 is smaller than the first threshold value T1, the recognition software determines that the high-brightness pixel region constitutes a valid Braille coding pattern 120a. Conversely, the recognition software recognizes that the high-brightness pixel region is only Noise. In this embodiment, the value of the first threshold T1 is 1/4, but the first threshold T1 can be determined by the actual value when the frame type mark is made, and the person familiar with the art can adjust it slightly according to the foregoing. The first judgment value C1 is calculated as the value of the first critical value T1, but it is still within the scope of the present invention.

Then, if the high-brightness pixels in the image form the Braille encoding pattern 120a, then the inside of the frame-shaped mark 122a is divided into 4×3 total 12 bits, and it is judged whether each bit is blank or Braille-encoded. Mark 124a. For example, firstly, whether there is a high-brightness mark in the bit is detected, the mark in the bit can be determined by defining an aspect ratio and an allowable ratio when the mark is made, and the aspect ratio is not easily affected by the change of the image distance, or The marking area is determined by a certain setting interval, the maximum value needs to be larger than the marking area of the closest image capturing distance, and the minimum value needs to be smaller than the marking area of the farthest image capturing distance, and an allowable ratio is set according to the length and width ratio of the marking, which meets the foregoing conditions. Then, the identification software will judge that the bit is the Braille encoding mark 124a, and vice versa, it is judged to be blank. In the present embodiment, the aspect ratio tolerance ratio is 0.3.

In other words, in the step of judging the frame mark 122a, the software can calculate the resolution of the Braille coding pattern 120a, and at the same time determine the distance between the body 112 and the Braille coding pattern 120a, so that the software can be inversely calculated. The area of the pixel that the Braille coded mark 124a should have. As described in the foregoing paragraph, the present invention can set the farthest image capturing distance and the closest image capturing distance as the allowable margin values to convert the pixel area corresponding conversion matrix, and the high brightness pixel area of the Braille encoding pattern is between the aforementioned pixels. Within the area, through the corresponding conversion matrix, the distance between the current mobile device and the Braille coding pattern can be further calculated from the area of the high-brightness reflective pixel, and the software can determine the location information of the current region and the more detailed portion of the mobile device.

In this way, the central control unit 114 can still parse the region identification information contained in the deformed Braille encoding pattern 120a to mb, thereby greatly improving the accuracy and success of the software identification. Since the present invention mainly performs positioning analysis by software analysis, the area identification system 100 of the present embodiment can greatly reduce the construction cost and facilitate the promotion to the home as compared with the conventional laser positioning or radio frequency identification technology. use.

It is to be noted that although the Braille encoding pattern 120 of the foregoing embodiment is a combination of the frame mark 122 and the Braille encoding mark 124, the present invention does not limit the structural composition of the Braille encoding pattern 120. For example, other embodiments may also change the frame-shaped mark 122 to a diagonal positioning mark or a four-corner positioning mark, which will be described below with reference to the embodiments.

3C-3F are front elevational views of four Braille encoding patterns in accordance with another embodiment of the present invention, and the areas of the Braille encoding indicia are aided by dashed lines for interpretation. Referring to FIGS. 3C-3F, the Braille encoding patterns 320a-320c of FIGS. 3C-3E respectively include two positioning marks 322a-322c and Braille encoding marks 324, wherein the positioning marks 322a-322c are located outside the area where the Braille encoding mark 324 is located. To define the area of the Braille encoded mark 324.

In detail, the two positioning marks 322a are located at the upper left corner and the lower right corner of the Braille encoding mark 324 region, and the two positioning marks 322b are located at the upper right corner and the lower left corner of the Braille encoding mark 324 region, and the two positioning marks 322c are Located above the Braille Encoded Mark 324 area and the left edge.

Similarly, when the central control unit 114 wants to determine whether the high-brightness pixels in the image constitute a text-encoding pattern, the outer edges of the positioning marks 322a-322c are first filled into a rectangle, and then the positioning marks 322a-322c are relatively rectangular. Whether the ratio of the area is within a certain range of values, and those skilled in the art can determine the specific range of values according to the foregoing description and actual design.

It should be noted that the present invention does not limit the shape and number of the positioning marks, nor does it limit the position of the positioning marks. For example, the positioning mark 322a is hook-shaped, and the shape of the positioning mark 322b is the same as that of the Braille coded mark 324, and the positioning mark 322c is elongated. In addition, although the number of the positioning marks 322a-322c is two, in other embodiments, three or more positioning marks may be used and arranged in each corner or edge of the Braille-encoded mark area. . Furthermore, the present invention may also employ different forms of positioning marks for identification of the positioning. As shown in FIG. 3F, the Braille encoding pattern 320d includes different types of positioning marks 322a-322c, and the positioning marks 322a-322c are respectively disposed at the lower left corner, the upper right corner, and the lower edge of the Braille encoding mark 324 region.

In the present embodiment, the Braille coding pattern 120 is disposed on the wall of the indoor space, and does not have a highly reflective effect on the fluorescent light source on the ceiling to avoid glare to the user of the indoor activity. Further, the light source generator 118 can generate the light source 118a at a specific angular range to illuminate the Braille encoding pattern 120. When the Braille encoding pattern 120 is illuminated by the light source 118a, the light source 118a is highly reflected back to the image receiver 116 to produce an image for analysis.

Referring to FIG. 1B again, in the embodiment, the central control unit 114 controls the main body 112, the image receiver 116, and the light source generator 118 in a wireless transmission manner, for example, as a computer. However, the present invention can also configure the central control unit 114 on the body 112 to directly manipulate the body 112, the image receiver 116, and the light source generator 118 in the direction of the connection. In addition, the present invention does not limit the types of the light source generator 118 and the image receiver 116. For example, the light source generator 118 can be a projection lamp, a light emitting diode, a halogen lamp, a fluorescent spotlight, or other suitable illumination source. The image receiver 116 can be a charge coupled device, a complementary MOS, a dome camera, or a digital camera.

In addition, although the aforementioned light sources are all illustrated by visible light, the present invention can also capture images by non-visible light. For example, the light source generator 118 can be an infrared light emitter to provide illumination of the infrared source to the Braille encoding pattern 120. Next, the Braille encoding pattern 120 can highly reflect the infrared source back to the image receiver 116, such as an infrared image receiver, to produce a light source. Under this architecture, the Braille coding pattern 120 can be designed not to reflect visible light, but can be hidden in a wall painting or other suitable location, and the Braille coding pattern 120 does not need to be designed for high reflection angles. .

In the foregoing embodiment, the Braille encoding pattern 120 may include at least two English letters or one English letter plus one digital combination of information amount as the area identification information, that is, 26×26=676 different representative meanings may represent the living room. , kitchen, dining room, children's room, charging station, computer room, study, guest room, corridor, etc., and the symbolic meaning of these areas can be defined by the user or the designer. Of course, the present invention also slightly adjusts the form of the Braille coding pattern depending on the amount of information. The following text will be further illustrated with a graphic description.

4A-4C are elevational views of a plurality of Braille encoding patterns in accordance with another embodiment of the present invention. Referring to FIGS. 4A-4C, the Braille encoding patterns 420a, 420b, 420c of the present embodiment are similar to the Braille encoding pattern 120 of FIG. 1C, and the Braille encoding patterns 420a, 420b, 420c also include frame-shaped marks 422a, 422b, 422c and Braille. The indicia 424a, 424b, 424c are encoded, but differ only in the form of the Braille encoding pattern.

Taking the Braille coding patterns 420a, 420c of FIGS. 4A and 4C as an example, they respectively have a single information amount combined with three English letters, and are represented as v and xyz, respectively. Taking the Braille coding pattern 420b of FIG. 4B as an example, the two Braille coding units are arranged up and down and used to denote pq. Those skilled in the art can easily understand this and will not repeat them here. It should be noted that the present invention does not limit the shape of the Braille coded mark. For example, the Braille coded marks 424a, 424b, and 424c may be square, triangular, and diamond-shaped, respectively, and the present invention does not limit the Braille coded pattern. The color of the box mark and the Braille coded mark.

It should be noted that although the aforementioned Braille encoding is mainly based on English encoding, the present invention does not limit the type of Braille encoding. For example, the Braille encoding of the present invention may also be Chinese encoding, Japanese encoding, Korean encoding or other. The right code. FIG. 5A is a local Braille encoding table of Chinese encoding mainly based on phonetic transcription, and FIG. 5B and FIG. 5C are partial Braille encoding tables of Japanese encoding and Korean encoding respectively, and those skilled in the art can easily modify Braille according to FIGS. 5A-5C. The meaning represented by the coding pattern is still within the scope of the present invention.

In summary, since the area recognition system of the present invention mainly uses the software to identify and analyze the image and thereby locate the position of the robot, the installation cost of the area recognition system is very low. In addition, the simple Braille encoding with the Braille encoding pattern of the highly reflective material can greatly reduce the difficulty of geometric recognition of image recognition, thereby greatly improving the accuracy and success of software recognition.

Although the present invention has been disclosed in the above embodiments, it is not intended to limit the present invention, and those skilled in the art can make some modifications and retouchings without departing from the spirit and scope of the present invention. The scope of protection is subject to the definition of the scope of the patent application attached.

100. . . Area recognition system

110. . . Mobile device

112. . . Ontology

114. . . Central control unit

116. . . Image receiver

118. . . Light source generator

118a. . . light source

120, 120a, 320a~320d, 420a~420c. . . Braille coding pattern

122, 122a, 422a~422c. . . Frame mark

124, 124a, 324, 424a~424c. . . Braille coded mark

322a~322c. . . Positioning mark

A. . . Area of pixels

L. . . length

W. . . width

1A is a block diagram of a region recognition system based on a mobile device in accordance with an embodiment of the present invention.

Figure 1B is a perspective view of the mobile device of Figure 1A.

1C and 1D are elevational views of the Braille encoding pattern of FIG. 1A.

Figure 2 is a Braille encoding table for English coding.

3A and 3B are schematic diagrams of images captured by the Braille encoding pattern of FIG. 1C at different angles.

3C-3F are elevational views of four Braille encoding patterns in accordance with another embodiment of the present invention.

4A-4C are elevational views of a plurality of Braille encoding patterns in accordance with another embodiment of the present invention.

Figures 5A-5C are partial Braille encoding tables encoded in Braille by country.

100. . . Area recognition system

110. . . Mobile device

112. . . Ontology

114. . . Central control unit

116. . . Image receiver

118. . . Light source generator

118a. . . light source

120. . . Braille coding pattern

Claims (23)

An area recognition system, comprising: a Braille coding pattern having an area identification information, wherein the material of the Braille coding pattern is a high reflective material, the Braille coding pattern comprises: a frame mark; and a Braille coding mark, configured in the a moving device, comprising: a body; a central control unit adapted to control the movement of the body; a light source generator disposed on the body, the light source generator being adapted to provide a light source, and the light source The Braille encoding pattern is adapted to reflect the light source; and an image receiver is disposed on the body and adapted to receive the image of the Braille encoding pattern and transmit to the central control unit, so that the central control unit identifies the information according to the area The position of the body is located. The area identification system of claim 1, wherein the Braille encoding of the Braille encoding pattern is Chinese encoding, English encoding, Japanese encoding, or Korean encoding. The area recognition system of claim 1, wherein the Braille coded mark has a shape of a circle, a square, a diamond or a triangle. The area recognition system of claim 1, wherein the Braille coding pattern reflects visible light or invisible light. The area recognition system of claim 1, wherein the high reflective material is a reflective particle, a reflective pigment or a reflective paint. The area recognition system of claim 4, wherein the invisible light is infrared light. The area recognition system of claim 1, wherein the light source generator is a projection lamp, a light emitting diode, a halogen lamp, a fluorescent lamp or a spotlight. The area recognition system of claim 1, wherein the image receiver is a charge coupled device, a complementary MOS, a dome camera or a digital camera. The area identification system of claim 1, wherein the central processing unit determines whether the received image is the Braille coding pattern or a noise based on a first determination value. The area identification system according to claim 9, wherein the central control unit fills the outer edge of the frame mark of the Braille coding pattern into a rectangle, and obtains the length L and the width W of the rectangle, and the central control unit calculates The pixel area A occupied by the frame mark is obtained by dividing the pixel area A occupied by the frame mark by the length L and dividing by the width W. An area recognition system includes: a Braille coding pattern having an area identification information, and the material of the Braille coding pattern is a high reflective material, the Braille coding pattern includes: two positioning marks, defining an area; and a Braille coding mark And disposed in the area defined by the positioning marks; a mobile device comprising: a body; a central control unit adapted to control the movement of the body; a light source generator disposed on the body, the light source The generator is adapted to provide a light source, and the Braille encoding pattern is adapted to reflect the light source; An image receiver is disposed on the body and adapted to receive the image of the Braille code pattern and transmit the image to the central control unit, so that the central control unit locates the position of the body according to the area identification information. The area recognition system of claim 11, wherein the Braille coded mark has a shape of a circle, a square, a diamond or a triangle. The area recognition system of claim 11, wherein the positioning marks are circular, hooked or elongated. The area identification system of claim 11, wherein the positioning marks are located at a corner or an edge of the area. The area identification system according to claim 11, wherein the Braille code of the Braille coding pattern is Chinese code, English code, Japanese code or Korean code. The area recognition system of claim 11, wherein the Braille coded mark has a shape of a circle, a square, a diamond or a triangle. The area recognition system of claim 11, wherein the Braille coding pattern reflects visible light or invisible light. The area recognition system of claim 11, wherein the high reflective material is a reflective particle, a reflective pigment or a reflective paint. The area recognition system of claim 17, wherein the invisible light is infrared light. The area recognition system of claim 11, wherein the light source generator is a projection lamp, a light emitting diode, a halogen lamp, a fluorescent lamp or a spotlight. The area recognition system of claim 11, wherein the image receiver is a charge coupled device, a complementary MOS, a ball Type camera or digital camera. The area identification system of claim 11, wherein the central processor is based on a first determination value to determine that the received image is the Braille coding pattern or a noise. The area identification system according to claim 22, wherein the central control unit fills the outer edge of the two positioning marks of the Braille coding pattern into a rectangle, and obtains the length L and the width W of the rectangle, and the central control unit calculates The pixel area A occupied by the frame mark is obtained by dividing the pixel area A occupied by the frame mark by the length L and dividing by the width W.