Many pressure transducers work by converting acoustic energy to electrical energy, and therefore could be said to be 'harvesting' the acoustic energy (though I would expect the process to be fairly inefficient). I would probably start by looking at piezo-type transducers - a quick Google search brought up some likely options.

As Craig has pointed out, all such transducers have very low efficiency. Piezo units can produce a higher voltage for a given sound pressure, compared to electrodynamic ones, but since the piezo devices have very high impedance, the current is minuscule, making the "harvested" power very low. Given more detail, I may be able to offe more information.

If possible, use a small, close parabolic or elliptical acoustically hard reflector, with line source at one focus, to redirect, boost, and capture the acoustic energy falling on a transducer --- a planar transducer for parabolic reflector, and a line transducer for an elliptical reflector (placed at second focal point of the ellipse. I'm assuming that the radius of the line source is small --- much smaller than its length.

Ronald

You may interest this paper "Estimation of directivity and sound power levels emitted by aircrafts during taxiing, for outdoor noise prediction purpose"

You may interest this paper "Estimation of directivity and sound power levels emitted by aircrafts during taxiing, for outdoor noise prediction purpose"

You may interest this paper "Estimation of directivity and sound power levels emitted by aircrafts during taxiing, for outdoor noise prediction purpose"

It depends on the nature of the line source? Is it by structural vibration radiation or something else

To make this question more clear, the line sound source I mentioned here is a laser beam that is periodically activated, which heat up the medium periodically (for example the air it pass through). The air illuminated by the laser then will produce a line sound source, but very weak in terms of sound intensity. 

Ronald Kessel mentioned a good way, but I  don't know if such line or planar transducers exist or not.

There is something called a "ribbon transducer" in acoustics, essentially a thin line microphone, designed mostly (I believe) for audio band frequencies. I know little about them. I think some sense acoustic pressure or vibration in air (which is probably what you want), and others sense more mechanical vibrations taken up at its base (clamped to body of a guitar, for guitar pick up, which is probably not what you want). 

Piezoelectric transducers can be manufactured in many different shapes and sizes. It may be possible to find or get a custom line-piezoelectric transducer made. Close attention to frequency bands of interest is required of course (presumably audio band or ultrasonic bands in your case).   

Ribbon transducers consist of a thin metal ribbon suspended in a magnetic field. They produce an output voltage due to induction in the ribbon moving in a magnetic field. the voltage is proportional to the particle velocity in  the sound wave. The output impedance is extremely low, so they are virtually always manufactured with a transformer in the same case. The frequency response extende from infrasonic to typically about 17,000 - 20,000 Hz. These are magnetic transducers, not piezo. Theie extremely low efficiency would seem to me to disqualify them from consideration for energy harvesting.

Hi Richard,

Thank you for your answer. I checked the principle of ribbon transducer today. I think it can be regarded as an small electric generator driven by sound. However as you said, the efficiency may limit its application. I suppose this is due to its low resonance frequency, which is under 100 Hz.

But it is an inspiring idea anyway, thank you!

Hi Ronald,

I checked the ribbon transducer, I agree it is kind of line transducer, but I don't think the "line" here has any relation to the sound source shape. It is just the shape of the transducer. 

But it is a good inspiration. Thank you!