视频中音频数据通过RGB image输出后的特征图片处理——音频处理2

第一步对得到的RGBimage进行剪切
下面展示一些 内联代码片。

from scipy import misc
import numpy as np
import cv2
import os
import pandas as pd

np.set_printoptions(threshold=np.inf)
def image():
    # print(os.path.abspath(image_path))
    # print(item)
    frame = cv2.imread(image_path+file) # 640x480x3 array
    cropped = frame[57:428,77:578] #对图片的白边进行剪切
    # cropped = frame[0:123, 0:501] #分成三等分，最上边的是high_frequency
    # cropped = frame[123:246, 0:501] #最中间的是middle_frequency
    # cropped = frame[247:371, 0:501] #最下边的是low_frequency
    cv2.imwrite(image_out+file,cropped)
image_path = 'E:/ffmpeg-latest-win64-static/2019_11_12_experimental_data/stimuli/feature/image_7_15/rgb/7_23/11/'
image_out = 'E:/ffmpeg-latest-win64-static/2019_11_12_experimental_data/stimuli/feature/image_7_15/cut/11/'
for file in os.listdir(image_path):
    print(file)
    image()

第二部将剪切好的high，middle，low图片中的RGB写为（0-255）的数字形式，并保存为csv

def image():

    frame = cv2.imread(image_path + file)  # 640x480x3 array
    rgb_frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)  # Convert to RGB
    print(cv2.mean(rgb_frame))
    r, g, b = cv2.split(rgb_frame)
    r_mean = r.mean()
    g_mean = g.mean()
    b_mean = b.mean()
    # cv2.imshow('bb',b)
    # cv2.imshow('gg',g)
    # cv2.imshow('rr',r)
    cv2.destroyAllWindows()
    return r_mean,g_mean,b_mean
image_path = 'E:/ffmpeg-latest-win64-static/2019_11_12_experimental_data/stimuli/feature/image_7_15/cut_rgb/low/11/'
image_out = 'E:/ffmpeg-latest-win64-static/2019_11_12_experimental_data/stimuli/feature/image_7_15/cut_rgb/low_csv/'
r_avg = [];
g_avg = [];
b_avg = []
for file in os.listdir(image_path):

    print(file)
    r_mean,g_mean,b_mean = image()
    r_avg.append(r_mean)
    g_avg.append(g_mean)
    b_avg.append(b_mean)
data = {'r': r_avg, 'g': g_avg, 'b': b_avg}
cols = ['r', 'g', 'b']
result = pd.DataFrame(data, columns=cols)
result.to_csv(image_out + '11_low.csv')