Academia.eduAcademia.edu

Figure from:Application of broadband dolphin mimetic sonar for discriminating target fish species

See full PDFdownloadDownload figure
of a small number of cycles, whereas the finless porpoise type has a narrower and width and is composed of a larger number of cycles. The term “dolphin” represents both dolphins and porpoises in his paper. Dolphins transmit highly focused sonar sounds forward from the head (Au 1993). It is well known that both the waveform and the frequency characteristics of the sounds measured off-axis are different from those measured on-axis. The typical on-axis sonar signals of the bottlenose dolphin and finless porpoise were recorded (Nakamura and Akamatsu 2004). Based on these sound characteristics, hree wave sinusoidal impulse was used as the alternative of the dolphin clicks because its mathematical expression is simple comparing with real biosonar signals at centroid frequency of 100 KHz. The -3 dB bandwidth of the tone burst was 30 kHz, from 84 to 114 kHz. It is clear that the transducer system can transmit and receive the majority of the spectral content of these broadband signals. The far-field distance was computed as 1.3 m at the highest frequency of 140 kHz, which was much shorter than the separation range. Because of the transducer frequency characteristics, the incident waveforms are distorted versions of the transmitting signals.

Figure 2 of a small number of cycles, whereas the finless porpoise type has a narrower and width and is composed of a larger number of cycles. The term “dolphin” represents both dolphins and porpoises in his paper. Dolphins transmit highly focused sonar sounds forward from the head (Au 1993). It is well known that both the waveform and the frequency characteristics of the sounds measured off-axis are different from those measured on-axis. The typical on-axis sonar signals of the bottlenose dolphin and finless porpoise were recorded (Nakamura and Akamatsu 2004). Based on these sound characteristics, hree wave sinusoidal impulse was used as the alternative of the dolphin clicks because its mathematical expression is simple comparing with real biosonar signals at centroid frequency of 100 KHz. The -3 dB bandwidth of the tone burst was 30 kHz, from 84 to 114 kHz. It is clear that the transducer system can transmit and receive the majority of the spectral content of these broadband signals. The far-field distance was computed as 1.3 m at the highest frequency of 140 kHz, which was much shorter than the separation range. Because of the transducer frequency characteristics, the incident waveforms are distorted versions of the transmitting signals.

Related Figures (6)

The broadband sound transmitting and receiving systems were constructed (Imaizumi et al. 2008) A custom made broadband transducer and power amplifier for transmitting and receiving was used (FURUNO electric Co. Ltd., Nishinomiya, Hyougo, Japan). Computer generated broadband signal was amplified and was sent to the transmit transducer. In the receiving system, the reflected wave was received by the same transducer, and the signal was amplified by the preamplifier. The output signals were observed, measured, and transformed into digital data by PXI system (National Instruments, USA) and the data were transferred to a personal computer harddisk. The transmitting sensitivity had a peak at 118 kHz, fell off rapidly below 60 kHz, and varied as much as 5 dB in the flat area between 67 and 134 kHz Fig. la. The beam width of the transmit transducer was 18.82° at 60 kHz and 9.03° at 120 kHz. The product of the transmitting and receiving sensitivities had a broadband characteristics of sensitivity product 70kHz to 140 kHz at a level of -10 dB.
The broadband sound transmitting and receiving systems were constructed (Imaizumi et al. 2008) A custom made broadband transducer and power amplifier for transmitting and receiving was used (FURUNO electric Co. Ltd., Nishinomiya, Hyougo, Japan). Computer generated broadband signal was amplified and was sent to the transmit transducer. In the receiving system, the reflected wave was received by the same transducer, and the signal was amplified by the preamplifier. The output signals were observed, measured, and transformed into digital data by PXI system (National Instruments, USA) and the data were transferred to a personal computer harddisk. The transmitting sensitivity had a peak at 118 kHz, fell off rapidly below 60 kHz, and varied as much as 5 dB in the flat area between 67 and 134 kHz Fig. la. The beam width of the transmit transducer was 18.82° at 60 kHz and 9.03° at 120 kHz. The product of the transmitting and receiving sensitivities had a broadband characteristics of sensitivity product 70kHz to 140 kHz at a level of -10 dB.
Table 1. Folk length, body depth and body width of three tuna species used for the measurement. C. Signal processing
Table 1. Folk length, body depth and body width of three tuna species used for the measurement. C. Signal processing
FIG3 Three net pens were fixed right next each other beside the pontoon. Each pen had single species of fish bigeye tuna, yellowfin tuna and skipjack tuna. They swam in circle and the acoustic beam was projected horizontally from the transducer as shown in blue long triangles. The position of the transducer was adjusted to point the beam to the center of each enclosure except for skipjack tuna. Because of the location of the skipjack pen, the acoustic beam axis was at an angle.
FIG3 Three net pens were fixed right next each other beside the pontoon. Each pen had single species of fish bigeye tuna, yellowfin tuna and skipjack tuna. They swam in circle and the acoustic beam was projected horizontally from the transducer as shown in blue long triangles. The position of the transducer was adjusted to point the beam to the center of each enclosure except for skipjack tuna. Because of the location of the skipjack pen, the acoustic beam axis was at an angle.
FIG4 Envelope extraction of the echo was conducted by cross correlation between incident and To estimate its temporal structure of a fish, the echo-envelope pattern was extracted. As shown by the solid curve of Fig. 4, this envelope pattern was calculated by the cross-correlation function between the incident wave and echo waveform (Oppenheim and Schafer, 1975). The temporal highlight structure was computed by extracting the local peak from the envelope pattern.
FIG4 Envelope extraction of the echo was conducted by cross correlation between incident and To estimate its temporal structure of a fish, the echo-envelope pattern was extracted. As shown by the solid curve of Fig. 4, this envelope pattern was calculated by the cross-correlation function between the incident wave and echo waveform (Oppenheim and Schafer, 1975). The temporal highlight structure was computed by extracting the local peak from the envelope pattern.
All of the three fish species swam in circle in the pen. The echogram showed sinusoidal pattern of the fish movement (FIG.4). Ordinate shows the distance from the transducer. The line appeared at 14 m is the nearer net of the enclosure from the transducer. The appeared at 20.5 m is the opposite side of the net. In between the net, which is the inside of the enclosure, periodical echo traces can be observed Since the fish swam in circle, strong echo came back when the fish pass through the acoustic beam perpendicular to the beam axis at 15 m and 19 m distance. Skipjack tuna appeared always to be approaching to the sound source. It is because the beam axis covered the half side of the enclosure unlike the case of other two species (FIG. 5). FIGS Echogram of bigeye tuna, yellowfin tuna, and skipjack tuna Each trace corresponds to individual fish except for skipjack tuna. Skipjack tuna swam in a group, which echogram showed a lil. RESULTS AND DISCUSSION
All of the three fish species swam in circle in the pen. The echogram showed sinusoidal pattern of the fish movement (FIG.4). Ordinate shows the distance from the transducer. The line appeared at 14 m is the nearer net of the enclosure from the transducer. The appeared at 20.5 m is the opposite side of the net. In between the net, which is the inside of the enclosure, periodical echo traces can be observed Since the fish swam in circle, strong echo came back when the fish pass through the acoustic beam perpendicular to the beam axis at 15 m and 19 m distance. Skipjack tuna appeared always to be approaching to the sound source. It is because the beam axis covered the half side of the enclosure unlike the case of other two species (FIG. 5). FIGS Echogram of bigeye tuna, yellowfin tuna, and skipjack tuna Each trace corresponds to individual fish except for skipjack tuna. Skipjack tuna swam in a group, which echogram showed a lil. RESULTS AND DISCUSSION
FIG. 6 Envelope (top) and echogram (bottom) patterns of three tuna species. Red center line of the envelope pattern indicate the swim bladder or spine, which are the major reflectors at the center of the fish body. There are two side robes above and below of the center. These side robes seemed to come from the both sides surface of the fish body. Intensity of the side robes were stronger for a yellowfin tuna and a skipjack tuna than that of a bigeye tuna.
FIG. 6 Envelope (top) and echogram (bottom) patterns of three tuna species. Red center line of the envelope pattern indicate the swim bladder or spine, which are the major reflectors at the center of the fish body. There are two side robes above and below of the center. These side robes seemed to come from the both sides surface of the fish body. Intensity of the side robes were stronger for a yellowfin tuna and a skipjack tuna than that of a bigeye tuna.