谷歌图片识别在线_图像识别库

全栈程序员站长

发布于 2022-09-21 11:03:59

4.7K0

发布于 2022-09-21 11:03:59

文章被收录于专栏：全栈程序员必看

大家好，又见面了，我是你们的朋友全栈君

一、准备模型

在这里，我们利用已经训练好的Googlenet进行物体图像的识别，进入Googlenet的GitHub地址，进入models文件夹，选择Googlenet

模型结构

在这里，我们利用之前讲到的网络模型绘制网站画出Googlenet的结构图如下：

在这里，pad就是给图像补零，pad:2就是补两圈零的意思；
LRN就是局部相应归一化，利用LRN可以提高模型识别的准确率；
Inception结构中，不同大小的卷积核意味着不同大小的感受野，最后的合并意味着不同尺度特征的融合。采用1,3,5为卷积核的大小，是因为使用步长为1，pad=0,1,2的方式采样之后得到的特征平面大小相同；
concat层用来合并数据，在这里合并的条件是数据的后三个参数要相同，所以在前面的inception结构中，我们使用了不同的卷积核大小和pad。

准备图片

在这里，我们找几张任意图片，然后放入Googlenet的文件夹下,，作为待识别的图片。

准备synset_words.txt文件

synset_words.txt是用来将物体的类别序号进行对应的文件，在识别过程中，我们先是得到序号，然后根据这个序号找到对应的物体种类。

使用python接口调用GoogleNet实现图像识别

在这里，我们用jupyter打开Googlenet.图像识别.ipynb文件,这里部分代码如下：

import caffe
import numpy as np
import matplotlib.pyplot as plt
import os
import PIL
from PIL import Image
import sys

#定义Caffe根目录 caffe_root = ‘E:/caffe-windows/’ #网络结构描述文件 deploy_file = caffe_root+‘models/bvlc_googlenet/deploy.prototxt’ #训练好的模型 model_file = caffe_root+‘models/bvlc_googlenet/bvlc_googlenet.caffemodel’

#cpu模式 caffe.set_mode_cpu()

#定义网络模型 net = caffe.Classifier(deploy_file, #调用deploy文件 model_file, #调用模型文件 mean=np.load(caffe_root +‘python/caffe/imagenet/ilsvrc_2012_mean.npy’).mean(1).mean(1), #调用均值文件 channel_swap=(2,1,0), #caffe中图片是BGR格式，而原始格式是RGB，所以要转化 raw_scale=255, #python中将图片存储为[0, 1]，而caffe中将图片存储为[0, 255]，所以需要一个转换 image_dims=(224, 224)) #输入模型的图片要是224*224的图片

#分类标签文件 imagenet_labels_filename = caffe_root +‘models/bvlc_googlenet/synset_words.txt’ #载入分类标签文件 labels = np.loadtxt(imagenet_labels_filename, str, delimiter=’\t’)

#对目标路径中的图像，遍历并分类 for root,dirs,files in os.walk(caffe_root+‘models/bvlc_googlenet/image/’): for file in files: #加载要分类的图片 image_file = os.path.join(root,file) input_image = caffe.io.load_image(image_file)

    <span class="token comment">#打印图片路径及名称</span>
    image_path <span class="token operator">=</span> os<span class="token punctuation">.</span>path<span class="token punctuation">.</span>join<span class="token punctuation">(</span>root<span class="token punctuation">,</span><span class="token builtin">file</span><span class="token punctuation">)</span>
    <span class="token keyword">print</span><span class="token punctuation">(</span>image_path<span class="token punctuation">)</span>
    
    <span class="token comment">#显示图片</span>
    img<span class="token operator">=</span>Image<span class="token punctuation">.</span><span class="token builtin">open</span><span class="token punctuation">(</span>image_path<span class="token punctuation">)</span>
    plt<span class="token punctuation">.</span>imshow<span class="token punctuation">(</span>img<span class="token punctuation">)</span>
    plt<span class="token punctuation">.</span>axis<span class="token punctuation">(</span><span class="token string">'off'</span><span class="token punctuation">)</span>
    plt<span class="token punctuation">.</span>show<span class="token punctuation">(</span><span class="token punctuation">)</span>
    
    <span class="token comment">#预测图片类别</span>
    prediction <span class="token operator">=</span> net<span class="token punctuation">.</span>predict<span class="token punctuation">(</span><span class="token punctuation">[</span>input_image<span class="token punctuation">]</span><span class="token punctuation">)</span>
    <span class="token keyword">print</span> <span class="token string">'predicted class:'</span><span class="token punctuation">,</span>prediction<span class="token punctuation">[</span><span class="token number">0</span><span class="token punctuation">]</span><span class="token punctuation">.</span>argmax<span class="token punctuation">(</span><span class="token punctuation">)</span>

    <span class="token comment"># 输出概率最大的前5个预测结果</span>
    top_k <span class="token operator">=</span> prediction<span class="token punctuation">[</span><span class="token number">0</span><span class="token punctuation">]</span><span class="token punctuation">.</span>argsort<span class="token punctuation">(</span><span class="token punctuation">)</span><span class="token punctuation">[</span><span class="token operator">-</span><span class="token number">5</span><span class="token punctuation">:</span><span class="token punctuation">]</span><span class="token punctuation">[</span><span class="token punctuation">:</span><span class="token punctuation">:</span><span class="token operator">-</span><span class="token number">1</span><span class="token punctuation">]</span>
    <span class="token keyword">for</span> node_id <span class="token keyword">in</span> top_k<span class="token punctuation">:</span>     
        <span class="token comment">#获取分类名称</span>
        human_string <span class="token operator">=</span> labels<span class="token punctuation">[</span>node_id<span class="token punctuation">]</span>
        <span class="token comment">#获取该分类的置信度</span>
        score <span class="token operator">=</span> prediction<span class="token punctuation">[</span><span class="token number">0</span><span class="token punctuation">]</span><span class="token punctuation">[</span>node_id<span class="token punctuation">]</span>
        <span class="token keyword">print</span><span class="token punctuation">(</span><span class="token string">'%s (score = %.5f)'</span> <span class="token operator">%</span> <span class="token punctuation">(</span>human_string<span class="token punctuation">,</span> score<span class="token punctuation">)</span><span class="token punctuation">)</span>

执行上述代码即可输出对物体种类的预测概率

最终得到的预测结果是根据可能性大小列出五个种类

        </div>
					<link href="https://csdnimg.cn/release/phoenix/mdeditor/markdown_views-2b43bc2447.css" rel="stylesheet">
            </div>
								<div class="hide-article-box text-center">
					<a class="btn" id="btn-readmore" data-track-view="{&quot;mod&quot;:&quot;popu_376&quot;,&quot;con&quot;:&quot;,https://blog.csdn.net/dagongsmallguy/article/details/83351025,&quot;}" data-track-click="{&quot;mod&quot;:&quot;popu_376&quot;,&quot;con&quot;:&quot;,https://blog.csdn.net/dagongsmallguy/article/details/83351025,&quot;}">阅读更多</a>
				</div>
				<script>
					(function(){
						function setArticleH(btnReadmore,posi){
							var winH = $(window).height();
							var articleBox = $("div.article_content");
							var artH = articleBox.height();
							if(artH > winH*posi){
								articleBox.css({
									'height':winH*posi+'px',
									'overflow':'hidden'
								})
								btnReadmore.click(function(){
									if(typeof window.localStorage === "object" && typeof window.csdn.anonymousUserLimit === "object"){
										if(!window.csdn.anonymousUserLimit.judgment()){
											window.csdn.anonymousUserLimit.Jumplogin();
											return false;
										}else if(!currentUserName){
											window.csdn.anonymousUserLimit.updata();
										}
									}
									
									articleBox.removeAttr("style");
									$(this).parent().remove();
								})
							}else{
								btnReadmore.parent().remove();
							}
						}
						var btnReadmore = $("#btn-readmore");
						if(btnReadmore.length>0){
							if(currentUserName){
								setArticleH(btnReadmore,3);
							}else{
								setArticleH(btnReadmore,1.2);
							}
						}
					})()
				</script>
				</article>

发布者：全栈程序员栈长，转载请注明出处：https://javaforall.cn/170449.html原文链接：https://javaforall.cn

本文参与腾讯云自媒体同步曝光计划，分享自作者个人站点/博客。

如有侵权请联系 cloudcommunity@tencent.com 删除

python

本文分享自作者个人站点/博客前往查看

如有侵权，请联系 cloudcommunity@tencent.com 删除。

本文参与腾讯云自媒体同步曝光计划，欢迎热爱写作的你一起参与！

python

登录后参与评论

0 条评论

热度