JPH0431160B2 - - Google Patents

Description

[Detailed description of the invention]

[Industrial Application Field] The present invention relates to an acoustic signal recognition device for a ship that automatically identifies whether an acoustic signal emitted from another vessel is a bell, a drum, or a whistle. [Prior Art] Conventionally, as one type of ship operation information, there are sound signals such as bells, gongs, and whistles emitted from other ships. By the way, depending on the combination of the bell, the drum, and the long and short sounds of the train whistle, the acoustic signal is as shown in Table 1 below.
It has various meanings as shown below.

【table】

[Table] Furthermore, the sound signals of bells, drums, and whistles have different tones, and based on the International Regulations for Preventing Collisions at Sea, the fundamental frequency of the sound signal of a train whistle is based on the length of the ship, that is, the ship's length L. It is determined as shown in Table 2 and FIG.

[Problem that the invention seeks to solve]

However, in the case of a super-automated ship with a small number of people, it is difficult to assign a crew member to monitor and listen to acoustic signals, and in this case, there is a problem that the acoustic signals cannot be identified by acquiring them as in the past. [Means for Solving the Problems] The present invention has a plurality of sound collecting devices each installed at a plurality of positions on a ship formed by a plurality of directional microphones, and each of the sound collecting devices collects sound. A sound source relative position measurement calculation unit calculates the relative position of the sound source of the other ship, and determines whether the sound source signal is the sound signal of a bell or gong by comparing the timbre of the sound source signal emitted by the sound source with the stored timbre. and determines whether the sound source signal is a whistle sound signal based on the frequency of the peak level of the sound signal, data on the relative position, and captain data of the other ship calculated from photographic data of the other ship. a sound source type determination unit that determines whether or not the other ship is present; and a movement detection unit that tracks the other ship and detects the movement of the other ship, and determines the presence or absence of an obstacle between the other ship and the main ship by radar exploration. and an obstacle discriminating section, the sound source signal discriminating section, the movement detecting and obstacle discriminating section, and the sound source signal based on the data and the sound source signal, and pre-stored bell, drum, and whistle patterns. An acoustic signal recognition device for a ship is characterized by comprising an acoustic signal recognition processing unit that identifies which pattern of an acoustic signal is a bell, a drum, or a whistle. [Function] Therefore, it is possible to automatically identify whether the sound signal emitted from another ship is a bell, drum, or whistle pattern signal, that is, which meaning of the signal in Table 1 is used. Ru. [Embodiment] Next, the present invention will be explained in detail with reference to FIGS. 1 to 4 showing one embodiment thereof. In FIG. 1, (1) is a sound source relative position measurement calculation unit, which has first and second sound collection devices 2, 3, a tape recorder 4, a background noise correction function 5, and a sound source relative position calculation function 6. 7 is a sound source type determination unit, which includes an image sensor 8, a ship captain calculation function 9, a timbre analysis function 10, a thermometer 11, a hygrometer 12, a memory function 13 for attenuation characteristics due to distance, temperature, and humidity, a bell bell, and a drive signal. It has a tone memory function 14 and a whistle tone memory function 15. 16 is a movement detection and obstacle discrimination unit, which has a sound source tracking function 17, a vessel absolute movement detection function 18,
It has a sound source absolute motion detection function 19 and a radar 20. 21 is an acoustic signal identification processing unit, which has a function 22 for storing patterns of bells, gongs, and whistles;
It has an acoustic signal identification function 23. The first and second sound collecting devices 2 and 3 are each formed by a plurality of directional microphones having the same characteristics, and when both the sound collecting devices 2 and 3 each have four directional microphones, as shown in FIG. The four directional microphones 2a, 2 of the sound collecting device 2
b, 2c, 2d are the rear center point O on the ship 24
The four directional microphones 3a, 3b, 3c, 3 of the sound collecting device 3 are installed at 90° intervals around the
d are placed around the front center point O' on the ship 24, spaced apart by 90 degrees. In addition, each microphone 2
The sound collection areas a to 2d and 3a to 3d are at the center point O,
Of course, it is set outside O'. There, another ship 26 having a sound source 25 of an acoustic signal
is located on the port side of the ship 24, for example, as shown in FIG.
The sound pressure levels of the microphones 3a and 3d of the sound collecting device 3 are high. That is, each microphone 2a~ of the sound collection device 2
The sound pressure level 2d and the sound pressure level of each of the microphones 3a to 3d of the sound collecting device 3 change depending on the relative position of the acoustic signal with respect to the ship 24. In the state shown in Fig. 2, the sound pressure level of the microphones 2a and 2b is ma, as shown in Fig. 3.
If mb, the direction of the sound source 25 with respect to the sound collector 2, that is, the angle α between the sound collector 2 and the sound source 25, is based on the ratio of the sound pressure levels ma and mb, and α=
It can be found from the formula tan ^-1 (ma/mb). Similarly, in the state shown in Fig. 2, if the sound pressure levels of microphones 3a and 3d are na and nd,
The direction of the sound source 25 with respect to the sound collecting device 3, that is, the angle β between the sound collecting device 3 and the sound source 25 is determined from the equation β=tan ⁻¹ (na/nd) based on the ratio of sound pressure levels. Once the angles α and β are determined, the relative position of the sound source 25 with respect to the ship 24 is determined using the cosine law. Therefore, as shown in FIG. 4a, the sound source relative position measurement calculation unit 1 constantly records the sound band signals collected by the sound collection devices 2 and 3 using the tape recorder 4, and 3, the signals collected at the same time are sequentially processed to calculate the angles α and β described above, and the relative position of the sound source 25 with respect to the ship 24 is repeatedly determined. Incidentally, the signals collected by the sound collectors 2 and 3 include unnecessary noise such as the sound of a ship's engine, wind noise, and the sound of waves, that is, signals of background noise. Therefore, the signal constantly recorded by the tape recorder 4 is input to the background noise correction function 5, and the correction function 5
In this way, the level of background noise that changes from time to time is measured and calculated, and the signal at the time of each relative position measurement and calculation is determined based on the background noise level immediately before or after calculation, for example, only the sound source signal emitted by a sound source such as the sound source 25. Correct the signal. Note that when there is an influence of wind, there is a possibility that the relative position of the sound source 25 may be measured or calculated deviating from the actual position. We also make corrections for the effects. Then, the signal corrected by the background noise correction function 5, that is, the sound collected during each relative position measurement and calculation, and the corrected sound source signal are input to the sound source relative position calculation function 6, and the calculation function 6 calculates the angle α , β are calculated, and the relative position of the sound source 25 with respect to the ship 24 is calculated using the cosine theorem. Next, in order to determine whether the sound source signal is an acoustic signal based on the tone color, the signal corrected by the background noise correction function 5 and the relative position data calculated by the sound source relative position calculation function 6 are used. , is input to the sound source type determining section 7. That is, the sound source signals that have been collected and subjected to background noise correction include signals other than the sound signals of bells, gongs, and train whistles. The bell, gong, and whistle each have a unique tone, and the fundamental frequency of the whistle differs depending on the captain L, as explained in Table 2 and FIG. 5. By the way, the sound pressure and timbre of the sound source signal change from the sound pressure and timbre of the sound source position based on attenuation due to the distance until reaching the ship 24, so for example, the timbre of a train whistle changes depending on the relative position of the sound source. Note that the amount of attenuation due to distance differs depending on the frequency, and this causes a change in timbre. Furthermore, the attenuation characteristics of the sound source signal also change depending on temperature and humidity. Therefore, in advance, the bell and gong tones are stored in the bell and gong tones memory function 14, and the train whistle tones of each frequency are stored in the whistle tones memory function 15, and the attenuation characteristic is memorized. 13 stores the attenuation characteristics for each frequency with distance, temperature, and humidity as variables. When the sound source signal and relative position data are input to the determination section 7, the timbre analysis function 10
The distance to the sound source is calculated based on the relative position data, etc., and as shown in FIG. Based on the stored data No. 14, the timbre of the input sound source signal is corrected by distance, temperature, and humidity. Furthermore, based on the comparison between the corrected tone of the sound source signal and the tone stored in the memory functions 14 and 15, the tone analysis function 10 determines whether the sound source signal is a bell bell,
It is determined whether or not it is a sound signal of a bell or a drum, and if it is a sound signal of a bell or a drum, the question "Is it a bell or a drum?"
to move to. When it is determined that the sound signal is not a bell or a drum, the timbre analysis function 10 next determines whether the sound source signal is a train whistle sound signal. By the way, since the sound signal of the whistle differs depending on the ship captain L, the image sensor 8 constantly photographs the direction of the sound source, and the ship captain calculation function 9 uses the data of the relative position and the photograph of the other ship 26 photographed by the image sensor 8. Based on data 26 other ships
Calculate the captain L of Note that the image sensor 8 is controlled by the sound source tracking function 17 or the like to always take an image in the direction of the sound source. Then, it was determined that the bell was not the sound signal of a gong.
When the question “Bell or gong?” in FIG. Frequency A is 70~130Hz, 130~200Hz, 200~250Hz,
It is determined whether the frequency belongs to any of the five frequency ranges of 250 to 350 Hz and 350 to 700 Hz. Further, when the frequency A belongs to any of the five frequency ranges, and the sound source signal is a whistle sound signal, the length of the ship length L must be as shown in Table 3 below.

[Table] When frequency A is determined to belong to one of the five frequency regions, the timbre analysis function 10 calculates the calculated length L for each frequency region in Table 3, as shown in Figure 4a. It is determined whether or not each of the captain L is satisfied, and based on the determination, it is determined whether or not the sound source signal is a sound signal of a train whistle. If it is a sound signal of a train whistle, the process moves to the step shown in FIG. 4b. . Incidentally, when it is determined that the sound signal is not a train whistle sound signal, the process moves to the step shown in FIG. 4b. When the timbre analysis function 10 of the determination unit 7 determines whether the sound source signal is an acoustic signal of a bell, a drum, or a train whistle, the acoustic signal identification processing unit 21 then selects one of the sound signals in Table 1. Identify what signals mean. That is, the pattern storage function 22 of the identification processing unit 21 stores patterns numbered from each number in Table 1 in advance. When the tone analysis function 10 determines that the sound signal is a bell or a drum, the processing unit 2
The identification function 23 of 1 refers to the memory pattern of the memory function 22 and starts processing from the point shown in FIG. If it is a pattern in which the number is heard for about 5 seconds, it is identified and recognized as an acoustic signal with a pattern indicating the meaning of the number. Further, if there is a pattern in which the bell is lit three times immediately before and after the bell sounds, the system identifies and recognizes that the sound signal has a pattern indicating the meaning of the number. Furthermore, even though the tone analysis function 10 determines that the signal is a bell or gong sound signal, if the pattern does not correspond to the pattern numbered in the pattern storage function 22, it is determined that the signal is not an audio signal. On the other hand, when the timbre analysis function 10 determines that the sound signal is a train whistle sound signal, the identification processing function 23 refers to the stored pattern in the pattern storage function 22 and starts processing from FIG. 4b. Then, the identification processing function 23 counts the number of whistles in the sound source signal, and also identifies the number of times of each long sound and short sound when a long sound and a short sound are combined based on the measurement of the length of the whistle. . However, when a train whistle consists of only long or short sounds, it is difficult to clearly distinguish between long sounds and short sounds. Therefore, the identification processing function 23 first determines the meaning of the sound signal of the train whistle based on the counted number of whistles, the number of identified long sounds and short sounds, and the memory pattern of the pattern storage function 22. and Table 4
Identify as shown.

〔Effect of the invention〕

As described above, according to the acoustic signal recognition device for a ship of the present invention, the sound source relative position measurement calculation unit 1 measures and calculates the relative position of the sound source of the other ship 26, and the sound source type determination unit 7 determines the sound source signal. , determines whether it is an acoustic signal of a drum or a whistle, and detects the movement of the other ship 26 by the movement detection and obstacle discrimination unit 16, and also detects the presence of an obstacle between the main ship 24 and the other ship 26. , none, and based on the data of the determining unit 7 and the determining unit 16 and the pre-stored bell, gong, and train whistle patterns, the sound signal identification processing unit 21 determines whether the sound source signal is a bell, gong, or whistle. By identifying which pattern it is, it is possible to automatically identify the meaning of the sound signal generated from the other ship 26, whether it is a bell, a gong, or a train whistle. This method is highly effective when applied to ships with a small number of people, such as super-automated ships.

[Brief explanation of drawings]

FIG. 1 is a block diagram of one embodiment of the acoustic signal recognition device for a ship according to the present invention, and FIG.
An explanatory diagram of the arrangement of the second sound collection device, Fig. 3 is an explanatory diagram for calculating the relative position, Fig. 4 a and b is a flowchart for explaining the operation of Fig. 1, and Fig. 5 is an explanation of the relationship between the acoustic signal and the captain. It is a diagram. 1... Sound source relative position measurement calculation unit, 7... Sound source type determination unit, 16... Movement detection and obstacle determination unit, 21
...Acoustic signal identification processing section.

Claims (1)

[Claims] 1 A sound source relative system that has a plurality of sound collection devices formed by a plurality of directional microphones and installed at a plurality of positions on the ship, and calculates the relative position of the sound source of another ship that is collected by each of the sound collection devices. A position measurement calculation unit determines whether the sound source signal is an acoustic signal of a bell or a gong by comparing the tone of the sound source signal emitted by the sound source with the stored tone, and
Determining whether the sound source signal is a whistle sound signal based on the frequency of the peak level of the sound source signal, data on the relative position, and captain data of the other ship calculated from photographic data of the other ship. a sound source type determination unit; a movement detection and obstacle determination unit that tracks the other ship and detects movement of the other ship, and determines the presence or absence of an obstacle between the other ship and the main ship by radar exploration; , the data of the sound source signal determination unit, the movement detection and obstacle determination unit, and the sound source signal, and a pre-stored bell bell;
Based on the pattern of the Dora and the whistle pattern and the pattern of the whistle, the sound of the ship that is characterized by the sound signal recognition process that identifies whether the sound signal is the acoustic signal of either the bell, the dora or whistle. Signal recognition device.