Animal sound detection method based on multi-band energy distribution in low signal-to-noise ratio environment
A technology with energy distribution and low signal-to-noise ratio, applied in speech analysis, instruments, etc., can solve problems such as detection effect limitations, and achieve good detection performance, low noise environment, and good robustness
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment Construction
[0050] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.
[0051] The existing methods for classification and detection of low signal-to-noise ratio sound events are as follows: figure 1 As shown in the thin line box in the lower part, that is, gray-scale logarithmic spectrogram, image feature extraction, and SVM classification. Using this method, the classification of sound events achieves better results. The detection rates of sound events with SNR of 20dB and 10dB are respectively 87.8% and 87.1%, especially in the case of 0dB, the detection rate reaches 74.4%. figure 1 As shown in the dotted line box in , the image feature extraction is through Jet mapping, and the gray logarithmic spectrum is mapped into three sub-images, and each sub-image is divided into 9×9 blocks, and then the mean value and variance of each block are extracted, that is, a total of 486 (2×3×9×9) dimensional vectors are used as featur...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


