Python列表刪除重復(fù)元素與圖像相似度判斷及刪除實(shí)例代碼

from img_similarity import runtwoImageSimilaryFun
import os
from PIL import Image
import shutil
import time
import numpy as np
 
def similar(path1, path2):
    img1 = Image.open(path1)
    img2 = Image.open(path2)
    w1 = img1.size[0] # 圖片的寬
    h1 = img2.size[1]  # 圖片的高
    w2 = img2.size[0] # 圖片的寬
    h2 = img2.size[1]  # 圖片的高
    w_err = abs(w1 - w2)/w1
    h_err = abs(h1 - h2)/h1
    if w_err > 0.1 or h_err >0.1:
        return 0
    else:
        phash, color_hist = runtwoImageSimilaryFun(path1, path2)
        if phash =8 or color_hist >=0.9:
            return 1
        else:
            return 0
 
 
path = './crop_img'
result_imgdirs_path = './removed_repeat_img'
folderlist = os.listdir(path)
folderlist.sort()
for item in folderlist:
    folder_path = path + '/' + item
    new_folder_path = result_imgdirs_path + '/' + item
    os.makedirs(new_folder_path)
 
    imglist = os.listdir(folder_path)
    imglist.sort()
 
    time_start = time.time()
 
    for i,item1 in enumerate(imglist):
        if item1 == '0':
            continue
        path1 = folder_path + '/' + item1
        for j, item2 in enumerate(imglist[i + 1:]):
            if item2 == '0':
                continue
            path2 = folder_path + '/' + item2
            t = similar(path1, path2)
            if t:
                #將判斷為相似的圖片在trans_list中的名字置‘0'，代表不需要復(fù)制
                imglist[i+j+1] = '0'
 
    imglist = list(set(imglist))
    imglist.remove('0')
 
    time_end = time.time()
    time_c = time_end - time_start
    print('{} similarity judgement list time cost {}s'.format(item, time_c))
 
 
    time_start = time.time()
    #移動(dòng)圖片
    for item3 in imglist:
        ori_img_path = folder_path + '/' + item3
        new_img_path = new_folder_path + '/' + item3
        shutil.copy(ori_img_path, new_img_path)
 
    time_end = time.time()
    time_c = time_end - time_start # 運(yùn)行所花時(shí)間
    print('{} move image time cost {}s'.format(item, time_c))

import cv2
import numpy as np
from PIL import Image
import requests
from io import BytesIO
import matplotlib
 
matplotlib.use('TkAgg')
import matplotlib.pyplot as plt
 
 
def aHash(img):
    # 均值哈希算法
    # 縮放為8*8
    img = cv2.resize(img, (8, 8))
    # 轉(zhuǎn)換為灰度圖
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    # s為像素和初值為0，hash_str為hash值初值為''
    s = 0
    hash_str = ''
    # 遍歷累加求像素和
    for i in range(8):
        for j in range(8):
            s = s + gray[i, j]
    # 求平均灰度
    avg = s / 64
    # 灰度大于平均值為1相反為0生成圖片的hash值
    for i in range(8):
        for j in range(8):
            if gray[i, j] > avg:
                hash_str = hash_str + '1'
            else:
                hash_str = hash_str + '0'
    return hash_str
 
 
def dHash(img):
    # 差值哈希算法
    # 縮放8*8
    img = cv2.resize(img, (9, 8))
    # 轉(zhuǎn)換灰度圖
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    hash_str = ''
    # 每行前一個(gè)像素大于后一個(gè)像素為1，相反為0，生成哈希
    for i in range(8):
        for j in range(8):
            if gray[i, j] > gray[i, j + 1]:
                hash_str = hash_str + '1'
            else:
                hash_str = hash_str + '0'
    return hash_str
 
 
def pHash(img):
    # 感知哈希算法
    # 縮放32*32
    img = cv2.resize(img, (32, 32))  # , interpolation=cv2.INTER_CUBIC
    # 轉(zhuǎn)換為灰度圖
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    # 將灰度圖轉(zhuǎn)為浮點(diǎn)型，再進(jìn)行dct變換
    dct = cv2.dct(np.float32(gray))
    # opencv實(shí)現(xiàn)的掩碼操作
    dct_roi = dct[0:8, 0:8]
 
    hash = []
    avreage = np.mean(dct_roi)
    for i in range(dct_roi.shape[0]):
        for j in range(dct_roi.shape[1]):
            if dct_roi[i, j] > avreage:
                hash.append(1)
            else:
                hash.append(0)
    return hash
 
 
def calculate(image1, image2):
    # 灰度直方圖算法
    # 計(jì)算單通道的直方圖的相似值
    hist1 = cv2.calcHist([image1], [0], None, [256], [0.0, 255.0])
    hist2 = cv2.calcHist([image2], [0], None, [256], [0.0, 255.0])
    # 計(jì)算直方圖的重合度
    degree = 0
    for i in range(len(hist1)):
        if hist1[i] != hist2[i]:
            degree = degree + \

                     (1 - abs(hist1[i] - hist2[i]) / max(hist1[i], hist2[i]))
        else:
            degree = degree + 1
    degree = degree / len(hist1)
    return degree
 
 
def classify_hist_with_split(image1, image2, size=(256, 256)):
    # RGB每個(gè)通道的直方圖相似度
    # 將圖像resize后，分離為RGB三個(gè)通道，再計(jì)算每個(gè)通道的相似值
    image1 = cv2.resize(image1, size)
    image2 = cv2.resize(image2, size)
    sub_image1 = cv2.split(image1)
    sub_image2 = cv2.split(image2)
    sub_data = 0
    for im1, im2 in zip(sub_image1, sub_image2):
        sub_data += calculate(im1, im2)
    sub_data = sub_data / 3
    return sub_data
 
 
def cmpHash(hash1, hash2):
    # Hash值對比
    # 算法中1和0順序組合起來的即是圖片的指紋hash。順序不固定，但是比較的時(shí)候必須是相同的順序。
    # 對比兩幅圖的指紋，計(jì)算漢明距離，即兩個(gè)64位的hash值有多少是不一樣的，不同的位數(shù)越小，圖片越相似
    # 漢明距離：一組二進(jìn)制數(shù)據(jù)變成另一組數(shù)據(jù)所需要的步驟，可以衡量兩圖的差異，漢明距離越小，則相似度越高。漢明距離為0，即兩張圖片完全一樣
    n = 0
    # hash長度不同則返回-1代表傳參出錯(cuò)
    if len(hash1) != len(hash2):
        return -1
    # 遍歷判斷
    for i in range(len(hash1)):
        # 不相等則n計(jì)數(shù)+1，n最終為相似度
        if hash1[i] != hash2[i]:
            n = n + 1
    return n
 
 
def getImageByUrl(url):
    # 根據(jù)圖片url 獲取圖片對象
    html = requests.get(url, verify=False)
    image = Image.open(BytesIO(html.content))
    return image
 
 
def PILImageToCV():
    # PIL Image轉(zhuǎn)換成OpenCV格式
    path = "/Users/waldenz/Documents/Work/doc/TestImages/t3.png"
    img = Image.open(path)
    plt.subplot(121)
    plt.imshow(img)
    print(isinstance(img, np.ndarray))
    img = cv2.cvtColor(np.asarray(img), cv2.COLOR_RGB2BGR)
    print(isinstance(img, np.ndarray))
    plt.subplot(122)
    plt.imshow(img)
    plt.show()
 
 
def CVImageToPIL():
    # OpenCV圖片轉(zhuǎn)換為PIL image
    path = "/Users/waldenz/Documents/Work/doc/TestImages/t3.png"
    img = cv2.imread(path)
    # cv2.imshow("OpenCV",img)
    plt.subplot(121)
    plt.imshow(img)
 
    img2 = Image.fromarray(cv2.cvtColor(img, cv2.COLOR_BGR2RGB))
    plt.subplot(122)
    plt.imshow(img2)
    plt.show()
 
 
def bytes_to_cvimage(filebytes):
    # 圖片字節(jié)流轉(zhuǎn)換為cv image
    image = Image.open(filebytes)
    img = cv2.cvtColor(np.asarray(image), cv2.COLOR_RGB2BGR)
    return img
 
 
def runAllImageSimilaryFun(para1, para2):
    # 均值、差值、感知哈希算法三種算法值越小，則越相似,相同圖片值為0
    # 三直方圖算法和單通道的直方圖 0-1之間，值越大，越相似。 相同圖片為1
    # t1,t2   14;19;10;  0.70;0.75
    # t1,t3   39 33 18   0.58 0.49
    # s1,s2  7 23 11     0.83 0.86  挺相似的圖片
    # c1,c2  11 29 17    0.30 0.31
 
    if para1.startswith("http"):
        # 根據(jù)鏈接下載圖片，并轉(zhuǎn)換為opencv格式
        img1 = getImageByUrl(para1)
        img1 = cv2.cvtColor(np.asarray(img1), cv2.COLOR_RGB2BGR)
 
        img2 = getImageByUrl(para2)
        img2 = cv2.cvtColor(np.asarray(img2), cv2.COLOR_RGB2BGR)
    else:
        # 通過imread方法直接讀取物理路徑
        img1 = cv2.imread(para1)
        img2 = cv2.imread(para2)
 
    hash1 = aHash(img1)
    hash2 = aHash(img2)
    n1 = cmpHash(hash1, hash2)
    print('均值哈希算法相似度aHash：', n1)
 
    hash1 = dHash(img1)
    hash2 = dHash(img2)
    n2 = cmpHash(hash1, hash2)
    print('差值哈希算法相似度dHash：', n2)
 
    hash1 = pHash(img1)
    hash2 = pHash(img2)
    n3 = cmpHash(hash1, hash2)
    print('感知哈希算法相似度pHash：', n3)
 
    n4 = classify_hist_with_split(img1, img2)
    print('三直方圖算法相似度：', n4)
 
    n5 = calculate(img1, img2)
    print("單通道的直方圖", n5)
    print("%d %d %d %.2f %.2f " % (n1, n2, n3, round(n4[0], 2), n5[0]))
    print("%.2f %.2f %.2f %.2f %.2f " % (1 - float(n1 / 64), 1 -
                                         float(n2 / 64), 1 - float(n3 / 64), round(n4[0], 2), n5[0]))
 
    plt.subplot(121)
    plt.imshow(Image.fromarray(cv2.cvtColor(img1, cv2.COLOR_BGR2RGB)))
    plt.subplot(122)
    plt.imshow(Image.fromarray(cv2.cvtColor(img2, cv2.COLOR_BGR2RGB)))
    plt.show()
 
 
def runtwoImageSimilaryFun(para1, para2):
    # 均值、差值、感知哈希算法三種算法值越小，則越相似,相同圖片值為0
    # 三直方圖算法和單通道的直方圖 0-1之間，值越大，越相似。 相同圖片為1
    # t1,t2   14;19;10;  0.70;0.75
    # t1,t3   39 33 18   0.58 0.49
    # s1,s2  7 23 11     0.83 0.86  挺相似的圖片
    # c1,c2  11 29 17    0.30 0.31
 
    if para1.startswith("http"):
        # 根據(jù)鏈接下載圖片，并轉(zhuǎn)換為opencv格式
        img1 = getImageByUrl(para1)
        img1 = cv2.cvtColor(np.asarray(img1), cv2.COLOR_RGB2BGR)
 
        img2 = getImageByUrl(para2)
        img2 = cv2.cvtColor(np.asarray(img2), cv2.COLOR_RGB2BGR)
    else:
        # 通過imread方法直接讀取物理路徑
        img1 = cv2.imread(para1)
        img2 = cv2.imread(para2)
 
 
    hash1 = pHash(img1)
    hash2 = pHash(img2)
    n3 = cmpHash(hash1, hash2)
 
    n4 = classify_hist_with_split(img1, img2)
 
    return n3, n4
 
 
 
if __name__ == "__main__":
    p1 = '/Users/Desktop/11/24.jpeg'
    p2 = '/Users/Desktop/11/25.jpeg'
    runAllImageSimilaryFun(p1, p2)