从嘈杂的音频python中检测2个频率

Question

我有一个AM无线电正在传输两个频率之一：1047hz或2093hz我试图录制音频并使用python来确定正在播放的频率（每个信号长1秒）。 问题是音频非常嘈杂，特别是在800-1000hz范围内，它干扰了我获得正确的信号。 如何滤除噪音？ 这是我的代码：

import sys, math
from aubio import source, pitch, freqtomidi




if len(sys.argv) < 2:
    print "Usage: %s <filename> [samplerate]" % sys.argv[0]
    sys.exit(1)

filename = sys.argv[1]
downsample = 1
samplerate = 44100 / downsample
if len( sys.argv ) > 2: samplerate = int(sys.argv[2])

win_s = 4096 / downsample # fft size
hop_s = 512  / downsample # hop size

s = source(filename, samplerate, hop_s)
samplerate = s.samplerate

tolerance = 0.8


pitch_o = pitch("yin", win_s, hop_s, samplerate)
pitch_o.set_tolerance(tolerance)

pitches = []
confidences = []
total_frames = 0
while True:
    samples, read = s()
    pitch = pitch_o(samples)[0]
    pitch = int(round(pitch))
    confidence = pitch_o.get_confidence()
    #if confidence < 0.6: pitch = 0.
    print "%f %f %f" % (total_frames / float(samplerate), pitch, confidence)
    pitches.append((total_frames/float(samplerate),pitch))
    confidences += [confidence]
    total_frames += read
    if read < hop_s: break

chunks = []
average_pitches = []

for i in range(len(pitches)):
    if pitches[i][1] < 950:
        del pitches[i]
    else:
        break

for i in reversed(range(len(pitches))):
    if pitches[i][1] == 0:
        del pitches[i]
    else:
        break


seconds = math.ceil(total_frames/float(samplerate))
start = pitches[0][0]



for i in range(int(pitches[len(pitches)-1][0]-start)+1):
    chunks.append([])


for pitch in pitches:
    if(pitch[0]==0):
        continue
    chunks[int(pitch[0]-start)].append(pitch[1])

for chunk in chunks:
    average_pitches.append(sum(chunk)/float(len(chunk)))

bits = []
for pitch in average_pitches:
    print pitch
    if 950 <= pitch < 1200:
        bits.append(0)
    if 1700 <= pitch <2100:
        bits.append(1) 

print bits