Speech coding system using variable threshold values for noise reduction

• US 4,918,734 A
• Filed: 05/21/1987
• Issued: 04/17/1990
• Est. Priority Date: 05/23/1986
• Status: Expired due to Fees

• Current Assignee: Hitachi, Ltd.
• Original Assignee: Hitachi, Ltd.
• Inventors: Kondo, Kazuhiro, Miyamoto, Takanori, Suzuki, Toshiro, Muramatsu, Ryujiro
• Primary Examiner(s): Harkcom, Gary V.
• Assistant Examiner(s): NOT, DEFINED
• Application Number:
  US07/052,395
• Time in Patent Office: 1,062 Days
• Field of Search: 381/46-47, 381/36-41, 381/110, 364/513.5
• US Class Current: 704/214
• CPC Class Codes: G10L 25/78 Detection of presence or absence of voice signals switching of direction of transmission by voice frequency in two-way loud-speaking telephone systems H04M9/10

Abstract

A speech coding system includes apparatus for generating a variable threshold dependent upon the power of an input speech signal, and a comparator for comparing the power of the input speech signal with the variable threshold value to generate a discriminating signal for discriminating between a period when a speech continues and a period when the speech pauses, to change the coding operation for the input speech signal in accordance with the level of the discriminating signal, thereby forming voiced and unvoiced frames independently of each other.

Claims

1. A speech coding system comprising:

means for calculating the power of an input speech signal at a predetermined time interval;

first attenuator means for attenuating the power of said input speech signal at a first attenuation rate and providing a first threshold value signal;

selector means, connected to receive said first threshold value signal and a second threshold value signal, for selecting one of said first and second threshold value signals having an amplitude larger than the other of said first and second threshold value signals and outputting said selected threshold value signal;

second attenuator means for attenuating the output of said selector means at a second attenuation rate smaller than said first attenuation rate

to produce said second threshold value signal, said second attenuator means including delay means for supplying said second threshold value signal to said selector means after a predetermined time delay;

comparator means for comparing the power of said input speech signal with the output of said selector means to generate a discriminating signal representative of whether the power of said input speech signal exceeds the output of said selector means; and

coding means for coding said input speech signal depending on said discriminating signal delivered from said comparator means to produce a voiced frame when said discriminating signal represents that the power of said input speech signal exceeds the output of said selector means and an unvoiced frame when said discriminating signal represents that the power of said input speech signal does not exceed the output of said selector means, said voiced frame including said coded input speech signal and an indicator for indicating that said input speech signal is a voice signal, and said unvoiced frame including said coded input speech signal and an indicator for indicating that said input speech signal is an unvoiced signal.

2. A speech coding system according to claim 1, wherein said system further comprises integrator means with leakage for integrating the power of said input speech signal and outputting an integrated power signal, wherein said first attenuator means produces said first threshold value signal from the output of said integrator means, and wherein said comparator means compares the output of said selector means with the output of said integrator means to generate said discriminating signal.