Sound directional sensing and enhancing method and systemTechnical Field
The invention relates to a technology in the field of audio processing, in particular to a sound directional sensing and enhancing method and system
Background
The existing voice enhancement method is mainly used for reducing the environmental noise by post-processing the voice signal through a signal processing algorithm and realizing the enhancement of the voice signal, and the method generally has a better suppression effect on the environmental noise mainly comprising Gaussian white noise, but has poor sound perception and enhancement effects in environments such as complex noise environments and noise aliasing, and meanwhile, the signal processing process is complex and time-consuming. While the directional voice sensing can be realized by the microphone array technology based on beam forming, the method often needs a huge microphone array, and meanwhile, the method is difficult to avoid being interfered by environmental noise and other sound source targets with relatively close distances due to low directional accuracy.
Disclosure of Invention
Aiming at the problems that the prior sound perception method is difficult to realize the voice capturing and enhancing under the conditions of strong background noise, clutter aliasing interference and the like, the invention provides a sound directional perception and enhancing method and system, which utilize a microwave transceiver and a microphone to perform multi-sensor fusion to realize the directional perception and sound enhancement of a sound source target.
The invention is realized by the following technical scheme:
The invention relates to a sound directional sensing and enhancing method, which is characterized in that micro-vibration information of a sound source to be detected is directionally collected through a microwave transceiver, simultaneously, an environment sound signal is synchronously collected through a microphone, and a sound signal extracted from the vibration information is fused with the environment sound signal, so that sound directional sensing and enhancing are realized.
The extraction refers to high-pass filtering processing is carried out on the vibration signal of the sound source target, and the sound signal of the directional sound source is inverted.
The fusion is to perform time-frequency analysis on the directional sound signal and the environment sound signal to obtain time-frequency distribution characteristics of the directional sound signal and the environment sound signal, divide the time-frequency distribution characteristics into time-frequency units with the length and the width of L, W, process the baseband signal of the microwave transceiver in each time-frequency unit, extract the vibration information of the target sound source as a basis to perform matched filtering on the environment sound signal, and then convert the filtering result from a time-frequency domain to a time-domain signal.
The microwave transceiver comprises a continuous wave microwave signal source, a power distributor, a power amplifier, a mixer, a low-pass filter, a conditioning circuit, a receiving antenna and a processor, wherein the continuous wave microwave signal source generates continuous wave microwave signals which are respectively output to the power amplifier through the power distributor and sent out by a transmitting antenna and output to the mixer to be mixed with reflected signals received by a receiving antenna, the mixer is connected with the low-pass filter and inputs the mixed signals into the low-pass filter, the low-pass filter is connected with the conditioning circuit, the conditioning circuit is connected with the processor, and the processor collects and processes baseband signals output by the conditioning circuit into directional vibration information of a sound source target.
Technical effects
The invention integrally solves the problem that the voice recognition and enhancement under the condition of strong background noise interference are difficult to realize in the existing voice sensing method, carries out matched filtering on the voice signals acquired by the microphone by utilizing the sound source target vibration information obtained by directional measurement of the microwave transceiver, and realizes the directional sensing and enhancement of the voice.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a schematic diagram of time-frequency characteristics of an embodiment;
FIG. 3 is a schematic diagram of an embodiment of a time-frequency block;
FIG. 4 is a schematic diagram showing the effect of the embodiment;
Fig. 5 is a schematic diagram of an embodiment application.
Detailed Description
As shown in fig. 1, this embodiment relates to a sound directional sensing and enhancing method, which includes the following steps:
Step 1, synchronously sensing voice signals in the environment by utilizing a microwave transceiver and a microphone, specifically, playing voice signals by utilizing one sound box, simultaneously playing voice interference signals by utilizing two sound boxes, simulating a test scene of complex noise interference, and synchronously acquiring the voice signals in the measurement environment by utilizing the microphone and the microwave transceiver. The antenna of the microwave transceiver faces to the sound source to be tested and is used for directing micro-vibration of the test sound source target, the microwave transceiver obtains a baseband signal by transmitting and receiving the microwave signal, and vibration displacement information of the sound source target is invertedAnd further, the vibration signal is subjected to high-pass filtering processing to convert the sound signal, wherein lambda is the wavelength of the microwave signal,Is the extracted interferometric phase evolution information.
Step 2, fusing the directional voice signal extracted by the microwave transceiver and the voice signal captured by the microphone to realize sound directional perception and enhancement, and specifically comprising the following steps:
And 2.1, performing time-frequency analysis on the sound signals perceived and inverted by the microwave transceiver to obtain time-frequency characteristics of the sound signals, as shown in fig. 2.
Step 2.2, as shown in fig. 3, after performing time-frequency analysis on the voice signal collected by the microphone to obtain time-frequency distribution, selecting window length L and window width W of a time-frequency unit, dividing the time-frequency distribution of the voice signal inverted by the microwave transceiver and the sound signal collected by the microphone into the same time-frequency units TFmicrowave (G, H) (G e [1, G ], H e [1, H ]) and TFmicrophone (G, H) (G e [1, G ], H e [1, H ]), wherein G and H are the number of rows and columns of the video unit formed after time-frequency division respectively, G and H are the row index and the column index of the selected time-frequency unit respectively, and TFmicrophone represents the time-frequency distribution;
And 2.3, processing the baseband signals of the microwave transceivers in each time-frequency unit, extracting the vibration information of the target sound source, and carrying out matched filtering on microphone voice signals, namely TFfusioo(g,h)＝TFmicrowave(g,h)*TFmicrophone (g, h), wherein the operation is matched filtering operation. As shown in fig. 4, the method can well separate the sound signal of the target sound source of interest from the output result of the matched filtering and noise reduction.
And 2.4, converting the matched filtering output result from a time-frequency domain to a time-domain signal.
As shown in FIG. 5, the sound orientation sensing and enhancing system for implementing the method according to the embodiment comprises a microwave transceiver, a microphone, a sound orientation sensing and enhancing signal processor and a sound result output unit, wherein the microwave transceiver is used for transmitting and receiving continuous wave microwave signals and then extracting vibration displacement information of a target surface from echo signals reflected by the target, the microphone is used for collecting sound signals in the environment, the sound orientation sensing and enhancing signal processor is used for processing the vibration displacement information of the target surface measured by the microwave transceiver and the sound signals collected by the microphone, and the sound result output unit is used for outputting the sound signals after orientation enhancement.
Compared with the prior art, the method realizes the directional perception of the sound and reduces the interference of other sound sources and environmental noise.
The foregoing embodiments may be partially modified in numerous ways by those skilled in the art without departing from the principles and spirit of the invention, the scope of which is defined in the claims and not by the foregoing embodiments, and all such implementations are within the scope of the invention.