Systems and methods for detection of a target sound
원문보기
IPC분류정보
국가/구분
United States(US) Patent
등록
국제특허분류(IPC7판)
G10L-025/51
H04R-029/00
H04R-001/32
H04L-029/08
출원번호
US-0909799
(2018-03-01)
등록번호
US-10236018
(2019-03-19)
발명자
/ 주소
Moore, Brian
Hoang, Viet
Brizel, Ken
Esfahany, Siamak Akhlaghi
Sandhu, Ajay
출원인 / 주소
Moore, Brian
대리인 / 주소
Bennett Jones LLP
인용정보
피인용 횟수 :
0인용 특허 :
10
초록▼
A system for detection of a target sound in an environment of a vehicle, includes an audio sensor, a computer processor, and a memory storing digital target sound templates produced by converting samples of the target sound in accordance with conversion parameters. The computer processor receives th
A system for detection of a target sound in an environment of a vehicle, includes an audio sensor, a computer processor, and a memory storing digital target sound templates produced by converting samples of the target sound in accordance with conversion parameters. The computer processor receives the sample sound signal from the audio sensor, converts the sample sound signal to a digital sample vector in accordance with the conversion parameters, compares the vector to each of the templates to determine if a degree of similarity of the vector to any one of the templates exceeds a predetermined threshold value, and notifies a user or a vehicle control system if the threshold value is exceeded.
대표청구항▼
1. A method for detection of a target sound, the method comprising the steps of: (a) receiving a sample sound signal generated by at least one audio sensor in response to a sample sound wave;(b) converting the sample sound signal to a digital sample vector in accordance with conversion parameters;(c
1. A method for detection of a target sound, the method comprising the steps of: (a) receiving a sample sound signal generated by at least one audio sensor in response to a sample sound wave;(b) converting the sample sound signal to a digital sample vector in accordance with conversion parameters;(c) comparing the sample vector to at least one template produced from the target sound using the same conversion parameters to determine if a degree of similarity of the vector to the at least one template exceeds a predetermined threshold value; and(d) generating an output signal if the threshold value is exceeded. 2. The method of claim 1, wherein the at least one audio sensor comprises at least two audio sensors, and the method comprises: (a) for each of the at least two audio sensors, performing steps (a) to (c) of claim 1, and determining a time index for an exceedance of the threshold value; and(b) determining a direction of the target sound based on a difference between the determined time indices. 3. The method of claim 1, wherein the target sound is a siren that sweeps between lower and upper frequency limits, and the templates represent data limited to either: (a) a narrow frequency band within the lower and upper frequency limits; or(b) at least one specific frequency within the narrow frequency band. 4. The method of claim 3, wherein the narrow frequency band is from about 600 Hz to about 1400 Hz. 5. The method of claim 3, wherein the at least one specific frequency comprises a plurality of specific frequencies, at least one, some or each of which is a prime number. 6. The method of claim 1, wherein the target sound is a siren that alternates between a lower frequency and an higher frequency, and the templates represent data limited to either: (a) a lower narrow frequency band centered on the lower frequency, and a higher narrow frequency band centered on the higher frequency; or(b) at least one specific frequency within each of the lower and higher narrow frequency bands. 7. The method of claim 6, wherein the lower and higher frequency bands provide a range from about 20% below to about 20% above the lower and higher frequencies, respectively. 8. The method of claim 1, further comprising the step of logarithmically amplifying the sample sound signal before being converted to the sample vector. 9. The method of claim 1 wherein the conversion parameter comprises an n-bit encoding of the sound signal, wherein n is less than or equal to 8. 10. The method of claim 9 wherein n is 1, 1.5 or 2. 11. The method of claim 10 wherein n is 1, and the conversion parameter is a voltage level crossing. 12. The method of claim 1, wherein one or both of the sample vector and the target sound template is less than about 10,000 bits long, or 5000 bits long, or 1000 bits long, or 100 bits long. 13. The method of claim 1, wherein steps (a) to (d) are performed on-board a vehicle in real-time, and the target sound is produced by a siren. 14. The method of any one of claim 1 wherein steps (b) and (c) involve only integer math. 15. A system for detection of a target sound, the system comprising: at least one audio sensor for generating a sample sound signal in response to a sample sound wave;a computer processor operatively connected to the at least one audio sensor; anda non-transitory computer readable memory operatively connected to the computer processor, and storing a plurality of digital target sound templates produced by converting a plurality of samples of the target sound in accordance with conversion parameters, and a set of instructions executable by the computer processor to implement the method of claim 1. 16. A continuous method of detecting a target sound in real time, comprising the steps of: (a) receiving sample sound signals generated by at least one audio sensor in response to sample sound waves;(b) logarithmically amplifying the sample sound signals and converting the amplified sample sound signals to a digital sample vector in accordance with conversion parameters;(c) comparing the sample vector to at least one digital target sound template produced from the target sound using the same conversion parameters to determine if a degree of similarity of the vector to the at least one template exceeds a predetermined threshold value; and(d) generating an output signal if the threshold value is exceeded. 17. The method of claim 16 wherein step (c) involves only integer math. 18. The method of claim 16 comprising the further step of notifying a vehicle user and/or a vehicle control system. 19. The method of claim 18 wherein the step of notifying a vehicle user comprises displaying an emergency vehicle location and optionally speed and direction of movement on a digital map. 20. A system for detection of a target sound, the system comprising: at least one audio sensor for generating a sample sound signal in response to a sample sound wave;a logarithmic amplifier operatively connected to the audio sensor for amplifying the sample sound signal;a computer processor operatively connected to the at least one audio sensor and the logarithmic amplifier; anda non-transitory computer readable memory operatively connected to the computer processor, and storing a plurality of digital target sound templates produced by converting a plurality of samples of the target sound in accordance with conversion parameters, and a set of instructions executable by the computer processor to implement the method of claim 16.
연구과제 타임라인
LOADING...
LOADING...
LOADING...
LOADING...
LOADING...
이 특허에 인용된 특허 (10)
St. John, Phyllis, Automobile emergency vehicle warning display system.
Ribnick Gerson D. (606 N. Larchmont Blvd. Los Angeles CA 90004) Gorenbein Steven M. (47 Alhambra Circle ; No. 2 Coral Gables FL 33134), Emergency vehicle warning system.
Hill James L. (11207 N. 39th St. Phoenix AZ 85028) Sloan John W. (2541 E. Yucca Phoenix AZ 85028), System for detecting the siren of an approaching emergency vehicle.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.