The signal-processing technology uses two microphone arrays and can isolate speakers with greater precision than directional microphones. The know-how could be used for teleconferences, exhibitions, manufacturing plants and other applications.
The software could also be used with microphones in mobile devices, giving smartphones the ability to refine sounds they pick up.
The approach, dubbed Area Sound Enhancement System, requires a minimum of two microphones. The mics are oriented at right angles to one another so that the areas they cover overlap.
By comparing the sound waves picked up by each mic, the signal-processing technology can isolate features common to each one while ignoring different features as noise. The result is an isolated sound, or speaker's voice, in a noisy environment, Oki said.
For instance, speakers in teleconferences whose sound is muddied by external noise would become easier to hear with two mics placed apart on a table in front of them.
It could also help improve speech-recognition rates for applications such as car navigation systems and automated dictation.
"This technology can isolate sound better than previous solutions such as directional microphones, which pick up all the sound in a certain direction," an Oki spokeswoman said, adding that the technique could work with a pair of smartphones that are synchronized while running the software.
The signal-processing algorithms don't require heavy-duty computation, so sound can be picked up in multiple areas with real-time results.
The Area Sound Enhancement System can also be used with a sound module that was developed by Oki in 2013. The module has four microphones and can help filter out unwanted noises in situations such as picking up sound when a smartphone is being used for a hands-free phone conversation.
Oki is still developing the Area Sound Enhancement System and has not decided on a timeline for commercialization yet, the spokeswoman said.
Tim Hornyak covers Japan and emerging technologies for The IDG News Service. Follow Tim on Twitter at @robotopia.