目錄

一.開源神經(jīng)網(wǎng)絡(AlexNet)

1.獲取數(shù)據(jù)集

2.神經(jīng)網(wǎng)絡模型

3.訓練神經(jīng)網(wǎng)絡

4.對模型進行預測

二、花卉識別系統(tǒng)搭建（flask）

1.構建頁面：

2.調用神經(jīng)網(wǎng)絡模型

3.系統(tǒng)識別結果

4.啟動系統(tǒng)：

三、總結

---

為什么會弄這個花卉識別系統(tǒng)？

這學期修了一門機器視覺的選修課，課設要是弄一個花卉識別的神經(jīng)網(wǎng)絡，所以我網(wǎng)上找了開源代碼進行了修改，最后成功跑起來，結果只有一個準確率（94%）

既然都跑了這個神經(jīng)網(wǎng)絡的代碼，那么干脆就把這個神經(jīng)網(wǎng)絡真正的使用起來，為此我繼續(xù)寫代碼，把這個神經(jīng)網(wǎng)絡弄成一個可視化界面（花卉識別系統(tǒng)）

一.開源神經(jīng)網(wǎng)絡(AlexNet)

1.獲取數(shù)據(jù)集

使用步驟如下：
* （1）在data_set文件夾下創(chuàng)建新文件夾"flower_data"
* （2）點擊鏈接下載花分類數(shù)據(jù)集http://download.tensorflow.org/example_images/flower_photos.tgz
* （3）解壓數(shù)據(jù)集到flower_data文件夾下
* （4）執(zhí)行"split_data.py"腳本自動將數(shù)據(jù)集劃分成訓練集train和驗證集val

split_data.py

import os
from shutil import copy, rmtree
import random


def mk_file(file_path: str):
    if os.path.exists(file_path):
        # 如果文件夾存在，則先刪除原文件夾在重新創(chuàng)建
        rmtree(file_path)
    os.makedirs(file_path)


def main():
    # 保證隨機可復現(xiàn)
    random.seed(0)

    # 將數(shù)據(jù)集中10%的數(shù)據(jù)劃分到驗證集中
    split_rate = 0.1

    # 指向你解壓后的flower_photos文件夾
    cwd = os.getcwd()
    data_root = os.path.join(cwd, "flower_data")
    origin_flower_path = os.path.join(data_root, "flower_photos")
    assert os.path.exists(origin_flower_path)
    flower_class = [cla for cla in os.listdir(origin_flower_path)
                    if os.path.isdir(os.path.join(origin_flower_path, cla))]

    # 建立保存訓練集的文件夾
    train_root = os.path.join(data_root, "train")
    mk_file(train_root)
    for cla in flower_class:
        # 建立每個類別對應的文件夾
        mk_file(os.path.join(train_root, cla))

    # 建立保存驗證集的文件夾
    val_root = os.path.join(data_root, "val")
    mk_file(val_root)
    for cla in flower_class:
        # 建立每個類別對應的文件夾
        mk_file(os.path.join(val_root, cla))

    for cla in flower_class:
        cla_path = os.path.join(origin_flower_path, cla)
        images = os.listdir(cla_path)
        num = len(images)
        # 隨機采樣驗證集的索引
        eval_index = random.sample(images, k=int(num*split_rate))
        for index, image in enumerate(images):
            if image in eval_index:
                # 將分配至驗證集中的文件復制到相應目錄
                image_path = os.path.join(cla_path, image)
                new_path = os.path.join(val_root, cla)
                copy(image_path, new_path)
            else:
                # 將分配至訓練集中的文件復制到相應目錄
                image_path = os.path.join(cla_path, image)
                new_path = os.path.join(train_root, cla)
                copy(image_path, new_path)
            print("[{}] processing [{}/{}]".format(cla, index+1, num), end="")  # processing bar
        print()

    print("processing done!")


if __name__ == '__main__':
    main()

2.神經(jīng)網(wǎng)絡模型

model.py

import torch.nn as nn
import torch

class AlexNet(nn.Module):
    def __init__(self, num_classes=1000, init_weights=False):
        super(AlexNet, self).__init__()
        # 用nn.Sequential()將網(wǎng)絡打包成一個模塊，精簡代碼
        self.features = nn.Sequential(   # 卷積層提取圖像特征
            nn.Conv2d(3, 48, kernel_size=11, stride=4, padding=2),  # input[3, 224, 224]  output[48, 55, 55]
            nn.ReLU(inplace=True), 									# 直接修改覆蓋原值，節(jié)省運算內(nèi)存
            nn.MaxPool2d(kernel_size=3, stride=2),                  # output[48, 27, 27]
            nn.Conv2d(48, 128, kernel_size=5, padding=2),           # output[128, 27, 27]
            nn.ReLU(inplace=True),
            nn.MaxPool2d(kernel_size=3, stride=2),                  # output[128, 13, 13]
            nn.Conv2d(128, 192, kernel_size=3, padding=1),          # output[192, 13, 13]
            nn.ReLU(inplace=True),
            nn.Conv2d(192, 192, kernel_size=3, padding=1),          # output[192, 13, 13]
            nn.ReLU(inplace=True),
            nn.Conv2d(192, 128, kernel_size=3, padding=1),          # output[128, 13, 13]
            nn.ReLU(inplace=True),
            nn.MaxPool2d(kernel_size=3, stride=2),                  # output[128, 6, 6]
        )
        self.classifier = nn.Sequential(   # 全連接層對圖像分類
            nn.Dropout(p=0.5),			   # Dropout 隨機失活神經(jīng)元，默認比例為0.5
            nn.Linear(128 * 6 * 6, 2048),
            nn.ReLU(inplace=True),
            nn.Dropout(p=0.5),
            nn.Linear(2048, 2048),
            nn.ReLU(inplace=True),
            nn.Linear(2048, num_classes),
        )
        if init_weights:
            self._initialize_weights()
            
	# 前向傳播過程
    def forward(self, x):
        x = self.features(x)
        x = torch.flatten(x, start_dim=1)	# 展平后再傳入全連接層
        x = self.classifier(x)
        return x
        
	# 網(wǎng)絡權重初始化，實際上 pytorch 在構建網(wǎng)絡時會自動初始化權重
    def _initialize_weights(self):
        for m in self.modules():
            if isinstance(m, nn.Conv2d):                            # 若是卷積層
                nn.init.kaiming_normal_(m.weight, mode='fan_out',   # 用（何）kaiming_normal_法初始化權重
                                        nonlinearity='relu')
                if m.bias is not None:
                    nn.init.constant_(m.bias, 0)                    # 初始化偏重為0
            elif isinstance(m, nn.Linear):            # 若是全連接層
                nn.init.normal_(m.weight, 0, 0.01)    # 正態(tài)分布初始化
                nn.init.constant_(m.bias, 0)          # 初始化偏重為0

3.訓練神經(jīng)網(wǎng)絡

train.py

# 導入包
import torch
import torch.nn as nn
from torchvision import transforms, datasets, utils
import matplotlib.pyplot as plt
import numpy as np
import torch.optim as optim
from model import AlexNet
import os
import json
import time

# 使用GPU訓練
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
with open(os.path.join("train.log"), "a") as log:
    log.write(str(device)+"
")

#數(shù)據(jù)預處理
data_transform = {
    "train": transforms.Compose([transforms.RandomResizedCrop(224),       # 隨機裁剪，再縮放成 224×224
                                 transforms.RandomHorizontalFlip(p=0.5),  # 水平方向隨機翻轉，概率為 0.5, 即一半的概率翻轉, 一半的概率不翻轉
                                 transforms.ToTensor(),
                                 transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))]),

    "val": transforms.Compose([transforms.Resize((224, 224)),  # cannot 224, must (224, 224)
                               transforms.ToTensor(),
                               transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))])}

#導入、加載 訓練集
# 導入訓練集
#train_set = torchvision.datasets.CIFAR10(root='./data', 	 # 數(shù)據(jù)集存放目錄
#										 train=True,		 # 表示是數(shù)據(jù)集中的訓練集
#                                        download=True,  	 # 第一次運行時為True，下載數(shù)據(jù)集，下載完成后改為False
#                                        transform=transform) # 預處理過程
# 加載訓練集                              
#train_loader = torch.utils.data.DataLoader(train_set, 	  # 導入的訓練集
#										   batch_size=50, # 每批訓練的樣本數(shù)
#                                          shuffle=False,  # 是否打亂訓練集
#                                          num_workers=0)  # num_workers在windows下設置為0


# 獲取圖像數(shù)據(jù)集的路徑
data_root = os.path.abspath(os.path.join(os.getcwd(), "../.."))  		# get data root path 返回上上層目錄
image_path = data_root + "/jqsj/data_set/flower_data/"  				 		# flower data_set path

# 導入訓練集并進行預處理
train_dataset = datasets.ImageFolder(root=image_path + "/train",		
                                     transform=data_transform["train"])
train_num = len(train_dataset)

# 按batch_size分批次加載訓練集
train_loader = torch.utils.data.DataLoader(train_dataset,	# 導入的訓練集
                                           batch_size=32, 	# 每批訓練的樣本數(shù)
                                           shuffle=True,	# 是否打亂訓練集
                                           num_workers=0)	# 使用線程數(shù)，在windows下設置為0

#導入、加載 驗證集
# 導入驗證集并進行預處理
validate_dataset = datasets.ImageFolder(root=image_path + "/val",
                                        transform=data_transform["val"])
val_num = len(validate_dataset)

# 加載驗證集
validate_loader = torch.utils.data.DataLoader(validate_dataset,	# 導入的驗證集
                                              batch_size=32, 
                                              shuffle=True,
                                              num_workers=0)

# 存儲 索引：標簽 的字典
# 字典，類別：索引 {'daisy':0, 'dandelion':1, 'roses':2, 'sunflower':3, 'tulips':4}
flower_list = train_dataset.class_to_idx
# 將 flower_list 中的 key 和 val 調換位置
cla_dict = dict((val, key) for key, val in flower_list.items())

# 將 cla_dict 寫入 json 文件中
json_str = json.dumps(cla_dict, indent=4)
with open('class_indices.json', 'w') as json_file:
    json_file.write(json_str)

#訓練過程
net = AlexNet(num_classes=5, init_weights=True)  	  # 實例化網(wǎng)絡（輸出類型為5，初始化權重）
net.to(device)									 	  # 分配網(wǎng)絡到指定的設備（GPU/CPU）訓練
loss_function = nn.CrossEntropyLoss()			 	  # 交叉熵損失
optimizer = optim.Adam(net.parameters(), lr=0.0002)	  # 優(yōu)化器（訓練參數(shù)，學習率）

save_path = './AlexNet.pth'
best_acc = 0.0

for epoch in range(150):
    ########################################## train ###############################################
    net.train()     					# 訓練過程中開啟 Dropout
    running_loss = 0.0					# 每個 epoch 都會對 running_loss  清零
    time_start = time.perf_counter()	# 對訓練一個 epoch 計時
    
    for step, data in enumerate(train_loader, start=0):  # 遍歷訓練集，step從0開始計算
        images, labels = data   # 獲取訓練集的圖像和標簽
        optimizer.zero_grad()	# 清除歷史梯度
        
        outputs = net(images.to(device))				 # 正向傳播
        loss = loss_function(outputs, labels.to(device)) # 計算損失
        loss.backward()								     # 反向傳播
        optimizer.step()								 # 優(yōu)化器更新參數(shù)
        running_loss += loss.item()
        
        # 打印訓練進度（使訓練過程可視化）
        rate = (step + 1) / len(train_loader)           # 當前進度 = 當前step / 訓練一輪epoch所需總step
        a = "*" * int(rate * 50)
        b = "." * int((1 - rate) * 50)
        with open(os.path.join("train.log"), "a") as log:
              log.write(str("train loss: {:^3.0f}%[{}->{}]{:.3f}".format(int(rate * 100), a, b, loss))+"
")
        print("train loss: {:^3.0f}%[{}->{}]{:.3f}".format(int(rate * 100), a, b, loss), end="")
    print()
    with open(os.path.join("train.log"), "a") as log:
              log.write(str('%f s' % (time.perf_counter()-time_start))+"
")
    print('%f s' % (time.perf_counter()-time_start))

    ########################################### validate ###########################################
    net.eval()    # 驗證過程中關閉 Dropout
    acc = 0.0  
    with torch.no_grad():
        for val_data in validate_loader:
            val_images, val_labels = val_data
            outputs = net(val_images.to(device))
            predict_y = torch.max(outputs, dim=1)[1]  # 以output中值最大位置對應的索引（標簽）作為預測輸出
            acc += (predict_y == val_labels.to(device)).sum().item()    
        val_accurate = acc / val_num
        
        # 保存準確率最高的那次網(wǎng)絡參數(shù)
        if val_accurate > best_acc:
            best_acc = val_accurate
            torch.save(net.state_dict(), save_path)
        with open(os.path.join("train.log"), "a") as log:
              log.write(str('[epoch %d] train_loss: %.3f  test_accuracy: %.3f 
' %
              (epoch + 1, running_loss / step, val_accurate))+"
")
        print('[epoch %d] train_loss: %.3f  test_accuracy: %.3f 
' %
              (epoch + 1, running_loss / step, val_accurate))
with open(os.path.join("train.log"), "a") as log:
      log.write(str('Finished Training')+"
")
print('Finished Training')

訓練結果后，準確率是94%

訓練日志如下：

4.對模型進行預測

predict.py

import torch
from model import AlexNet
from PIL import Image
from torchvision import transforms
import matplotlib.pyplot as plt
import json

# 預處理
data_transform = transforms.Compose(
    [transforms.Resize((224, 224)),
     transforms.ToTensor(),
     transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))])

# load image
img = Image.open("pgy2.jpg")
#plt.imshow(img)
# [N, C, H, W]
img = data_transform(img)
# expand batch dimension
img = torch.unsqueeze(img, dim=0)

# read class_indict
try:
    json_file = open('./class_indices.json', 'r')
    class_indict = json.load(json_file)
except Exception as e:
    print(e)
    exit(-1)

# create model
model = AlexNet(num_classes=5)
# load model weights
model_weight_path = "./AlexNet.pth"
#, map_location='cpu'
model.load_state_dict(torch.load(model_weight_path, map_location='cpu'))

# 關閉 Dropout
model.eval()
with torch.no_grad():
    # predict class
    output = torch.squeeze(model(img))     # 將輸出壓縮，即壓縮掉 batch 這個維度
    predict = torch.softmax(output, dim=0)
    predict_cla = torch.argmax(predict).numpy()
print(class_indict[str(predict_cla)], predict[predict_cla].item())
plt.show()

接著對其中一個花卉圖片進行識別，其結果如下：

可以看到只有一個識別結果（daisy雛菊）和準確率1.0是100%（范圍是0~1，所以1對應100%）

為了方便使用這個神經(jīng)網(wǎng)絡，接著我們將其開發(fā)成一個可視化的界面操作

二、花卉識別系統(tǒng)搭建（flask）

1.構建頁面：

upload.html

<!DOCTYPE html>
<html lang="en">
<head>
    <meta charset="UTF-8">
    <title>李運辰-花卉識別系統(tǒng)v1.0</title>
	<link rel="stylesheet" type="text/css" href="../static/css/bootstrap.min.css">
	<link rel="stylesheet" type="text/css" href="../static/css/fileinput.css">
    <script src="../static/js/jquery-2.1.4.min.js"></script>
	
    <script src="../static/js/bootstrap.min.js"></script> 
	<script src="../static/js/fileinput.js"></script> 
	<script src="../static/js/locales/zh.js"></script> 
	
</head>
<body>
    <h1   align="center">李運辰-花卉識別系統(tǒng)v1.0</h1>
	<div align="center">
    <form action="" enctype='multipart/form-data' method='POST'>
        <input type="file" name="file"  class="file"  data-show-preview="false"/>
        <br>
        <input type="submit" value="上傳" class="btn btn-primary button-new "/>
    </form>
	</div>
</body>
</html>

2.調用神經(jīng)網(wǎng)絡模型

main.py

# coding:utf-8

from flask import Flask, render_template, request, redirect, url_for, make_response, jsonify
from werkzeug.utils import secure_filename
import os
import time


###################
#模型所需庫包
import torch
from model import AlexNet
from PIL import Image
from torchvision import transforms
import matplotlib.pyplot as plt
import json

# read class_indict
try:
    json_file = open('./class_indices.json', 'r')
    class_indict = json.load(json_file)
except Exception as e:
    print(e)
    exit(-1)

# create model
model = AlexNet(num_classes=5)
# load model weights
model_weight_path = "./AlexNet.pth"
#, map_location='cpu'
model.load_state_dict(torch.load(model_weight_path, map_location='cpu'))

# 關閉 Dropout
model.eval()

###################
from datetime import timedelta
# 設置允許的文件格式
ALLOWED_EXTENSIONS = set(['png', 'jpg', 'JPG', 'PNG', 'bmp'])

def allowed_file(filename):
    return '.' in filename and filename.rsplit('.', 1)[1] in ALLOWED_EXTENSIONS

app = Flask(__name__)
# 設置靜態(tài)文件緩存過期時間
app.send_file_max_age_default = timedelta(seconds=1)

#圖片裝換操作
def tran(img_path):
     # 預處理
    data_transform = transforms.Compose(
        [transforms.Resize((224, 224)),
         transforms.ToTensor(),
         transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))])

    # load image
    img = Image.open("pgy2.jpg")
    #plt.imshow(img)
    # [N, C, H, W]
    img = data_transform(img)
    # expand batch dimension
    img = torch.unsqueeze(img, dim=0)
    return img
    

@app.route('/upload', methods=['POST', 'GET'])  # 添加路由
def upload():
    path=""
    if request.method == 'POST':
        f = request.files['file']
        if not (f and allowed_file(f.filename)):
            return jsonify({"error": 1001, "msg": "請檢查上傳的圖片類型，僅限于png、PNG、jpg、JPG、bmp"})

        basepath = os.path.dirname(__file__)  # 當前文件所在路徑
        path = secure_filename(f.filename)
        upload_path = os.path.join(basepath, 'static/images', secure_filename(f.filename))  # 注意：沒有的文件夾一定要先創(chuàng)建，不然會提示沒有該路徑
        # upload_path = os.path.join(basepath, 'static/images','test.jpg')  #注意：沒有的文件夾一定要先創(chuàng)建，不然會提示沒有該路徑
        print(path)

        img = tran('static/images'+path)
        ##########################
        #預測圖片
        with torch.no_grad():
            # predict class
            output = torch.squeeze(model(img))     # 將輸出壓縮，即壓縮掉 batch 這個維度
            predict = torch.softmax(output, dim=0)
            predict_cla = torch.argmax(predict).numpy()
            res = class_indict[str(predict_cla)]
            pred = predict[predict_cla].item()
            #print(class_indict[str(predict_cla)], predict[predict_cla].item())
        res_chinese = ""
        if res=="daisy":
            res_chinese="雛菊"
        if res=="dandelion":
            res_chinese="蒲公英"
        if res=="roses":
            res_chinese="玫瑰"
        if res=="sunflower":
            res_chinese="向日葵"
        if res=="tulips":
            res_chinese="郁金香"

        #print('result:', class_indict[str(predict_class)], 'accuracy:', prediction[predict_class])
        ##########################
        f.save(upload_path)
        pred = pred*100
        return render_template('upload_ok.html', path=path, res_chinese=res_chinese,pred = pred, val1=time.time())

    return render_template('upload.html')

if __name__ == '__main__':
    # app.debug = True
    app.run(host='127.0.0.1', port=80,debug = True)

3.系統(tǒng)識別結果

<!DOCTYPE html>
<html lang="en">
<head>
    <meta charset="UTF-8">
    <title>李運辰-花卉識別系統(tǒng)v1.0</title>
	<link rel="stylesheet" type="text/css" href="../static/css/bootstrap.min.css">
	<link rel="stylesheet" type="text/css" href="../static/css/fileinput.css">
    <script src="../static/js/jquery-2.1.4.min.js"></script>
	
    <script src="../static/js/bootstrap.min.js"></script> 
	<script src="../static/js/fileinput.js"></script> 
	<script src="../static/js/locales/zh.js"></script> 
</head>
<body>
    <h1  align="center">李運辰-花卉識別系統(tǒng)v1.0</h1>
	<div align="center">
    <form action="" enctype='multipart/form-data' method='POST'>
        <input type="file" name="file" class="file"  data-show-preview="false"/>
        <br>
        <input type="submit" value="上傳" class="button-new btn btn-primary"/>
    </form>
	<p >識別結果:{{res_chinese}}</p>
	</br>
	<p >準確率:{{pred}}%</p>
    <img src="{{%20'./static/images/'+path%20}}" width="400" height="400" alt=""/>
	</div>
</body>
</html>

4.啟動系統(tǒng)：

python main.py

接著在瀏覽器在瀏覽器里面訪問

http://127.0.0.1/upload

出現(xiàn)如下界面：

最后來一個識別過程的動圖

三、總結

ok，這個花卉系統(tǒng)就已經(jīng)搭建完成了，是不是超級簡單，我也是趁著修了這個機器視覺這么課，才弄這么一個系統(tǒng)，回顧一下之前的知識，哈哈哈。

如果有任何問題歡迎在下方評論

最后附上系統(tǒng)完整源碼獲取：公眾號回復：花識別系統(tǒng)

【加群獲取學習資料QQ群：901381280】

【各種爬蟲源碼獲取方式】

識別文末二維碼，回復：爬蟲源碼

歡迎關注公眾號：Python爬蟲數(shù)據(jù)分析挖掘，方便及時閱讀最新文章

回復【開源源碼】免費獲取更多開源項目源碼；

總結

以上是生活随笔為你收集整理的教你搭建一个花卉识别系统(超级简单)的全部內(nèi)容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網(wǎng)站內(nèi)容還不錯，歡迎將生活随笔推薦給好友。