devbox
代码探险家
这个屌丝很懒，什么也没留下！
关注作者
热门标签
热门文章
当前位置:   article > 正文

C# Onnx Chinese CLIP 通过一句话从图库中搜出来符合要求的图片_c# winform onnx 识别图片

作者：代码探险家 | 2024-07-01 08:00:30

赞

踩

c# winform onnx 识别图片

目录

效果

生成图片特征

查找踢足球的人

测试图片

模型信息

image_model.onnx

text_model.onnx

项目

代码

Form1.cs

Clip.cs

下载

C# Onnx Chinese CLIP 通过一句话从图库中搜出来符合要求的图片

效果

生成图片特征

查找踢足球的人

测试图片

模型信息

image_model.onnx

Inputs
-------------------------
name：image
tensor：Float[1, 3, 224, 224]
---------------------------------------------------------------

Outputs
-------------------------
name：unnorm_image_features
tensor：Float[1, 512]
---------------------------------------------------------------

text_model.onnx

Inputs
-------------------------
name：text
tensor：Int64[1, 52]
---------------------------------------------------------------

Outputs
-------------------------
name：unnorm_text_features
tensor：Float[1, 512]
---------------------------------------------------------------

项目

代码

Form1.cs


using System;
using System.Collections.Generic;
using System.Drawing;
using System.IO;
using System.Linq;
using System.Text;
using System.Windows.Forms;

namespace Onnx_Demo
{
    public partial class Form1 : Form
    {
        public Form1()
        {
            InitializeComponent();
        }

        Clip mynet = new Clip("model/image_model.onnx", "model/text_model.onnx", "model/myvocab.txt");

        float[] imagedir_features;
        string image_dir = "test_img";
        StringBuilder sb = new StringBuilder();

        private void button2_Click(object sender, EventArgs e)
        {
            //特征向量 可以存二进制文件或者向量数据库
            imagedir_features = mynet.generate_imagedir_features(image_dir);
            txtInfo.Text = "生成完成！";
            txtInfo.Text += "有" + mynet.imgnum + "张图片,特征向量长度=" + imagedir_features.Length;
        }

        private void button3_Click(object sender, EventArgs e)
        {
            if (imagedir_features == null)
            {
                MessageBox.Show("请先生成图片特征！");
                return;
            }

            sb.Clear();
            txtInfo.Text = "";
            lblInfo.Text = "";
            pictureBox1.Image = null;

            string input_text = txt_input_text.Text;
            if (string.IsNullOrEmpty(input_text))
            {
                return;
            }
            List<Dictionary<string, float>> top5imglist = mynet.input_text_search_image(input_text, imagedir_features, mynet.imglist);

            sb.AppendLine("top5:");
            foreach (var item in top5imglist)
            {
                sb.AppendLine(Path.GetFileName(item.Keys.First()) + "  相似度:" + item[item.Keys.First()].ToString("F2"));
            }

            txtInfo.Text = sb.ToString();
            lblInfo.Text = Path.GetFileName(top5imglist[0].Keys.First());
            pictureBox1.Image = new Bitmap(top5imglist[0].Keys.First());

        }

        private void Form1_Load(object sender, EventArgs e)
        {

        }
    }
}

  1.  
  2. using System;
  3. using System.Collections.Generic;
  4. using System.Drawing;
  5. using System.IO;
  6. using System.Linq;
  7. using System.Text;
  8. using System.Windows.Forms;
  9.  
  10. namespace Onnx_Demo
  11. {
  12.     public partial class Form1 : Form
  13.     {
  14.         public Form1()
  15.         {
  16.             InitializeComponent();
  17.         }
  18.  
  19.         Clip mynet = new Clip("model/image_model.onnx", "model/text_model.onnx", "model/myvocab.txt");
  20.  
  21.         float[] imagedir_features;
  22.         string image_dir = "test_img";
  23.         StringBuilder sb = new StringBuilder();
  24.  
  25.         private void button2_Click(object sender, EventArgs e)
  26.         {
  27.             //特征向量 可以存二进制文件或者向量数据库
  28.             imagedir_features = mynet.generate_imagedir_features(image_dir);
  29.             txtInfo.Text = "生成完成！";
  30.             txtInfo.Text += "有" + mynet.imgnum + "张图片,特征向量长度=" + imagedir_features.Length;
  31.         }
  32.  
  33.         private void button3_Click(object sender, EventArgs e)
  34.         {
  35.             if (imagedir_features == null)
  36.             {
  37.                 MessageBox.Show("请先生成图片特征！");
  38.                 return;
  39.             }
  40.  
  41.             sb.Clear();
  42.             txtInfo.Text = "";
  43.             lblInfo.Text = "";
  44.             pictureBox1.Image = null;
  45.  
  46.             string input_text = txt_input_text.Text;
  47.             if (string.IsNullOrEmpty(input_text))
  48.             {
  49.                 return;
  50.             }
  51.             List<Dictionary<string, float>> top5imglist = mynet.input_text_search_image(input_text, imagedir_features, mynet.imglist);
  53.             sb.AppendLine("top5:");
  54.             foreach (var item in top5imglist)
  55.             {
  56.                 sb.AppendLine(Path.GetFileName(item.Keys.First()) + "  相似度:" + item[item.Keys.First()].ToString("F2"));
  57.             }
  58.  
  59.             txtInfo.Text = sb.ToString();
  60.             lblInfo.Text = Path.GetFileName(top5imglist[0].Keys.First());
  61.             pictureBox1.Image = new Bitmap(top5imglist[0].Keys.First());
  62.  
  63.         }
  64.  
  65.         private void Form1_Load(object sender, EventArgs e)
  66.         {
  67.  
  68.         }
  69.     }
  70. }

Clip.cs

public class Clip
    {
        int inpWidth = 224;
        int inpHeight = 224;
        float[] mean = new float[] { 0.48145466f, 0.4578275f, 0.40821073f };
        float[] std = new float[] { 0.26862954f, 0.26130258f, 0.27577711f };

        int context_length = 52;
        int len_text_feature = 512;

        Net net;
        float[] image_features_input;

        SessionOptions options;
        InferenceSession onnx_session;
        Tensor<long> input_tensor;
        List<NamedOnnxValue> input_container;
        IDisposableReadOnlyCollection<DisposableNamedOnnxValue> result_infer;
        DisposableNamedOnnxValue[] results_onnxvalue;
        Tensor<float> result_tensors;

        TokenizerBase tokenizer;

        int[] text_tokens_input;
        float[,] text_features_input;

        public int imgnum = 0;
        public List<string> imglist = new List<string>();

        public Clip(string image_modelpath, string text_modelpath, string vocab_path)
        {
            net = CvDnn.ReadNetFromOnnx(image_modelpath);

            // 创建输出会话，用于输出模型读取信息
            options = new SessionOptions();
            options.LogSeverityLevel = OrtLoggingLevel.ORT_LOGGING_LEVEL_INFO;
            options.AppendExecutionProvider_CPU(0);// 设置为CPU上运行
            // 创建推理模型类，读取本地模型文件
            onnx_session = new InferenceSession(text_modelpath, options);//model_path 为onnx模型文件的路径
            // 创建输入容器
            input_container = new List<NamedOnnxValue>();

            load_tokenizer(vocab_path);

        }

        void load_tokenizer(string vocab_path)
        {

            tokenizer = new TokenizerClipChinese();
            tokenizer.load_tokenize(vocab_path);
            text_tokens_input = new int[1024 * context_length];
        }

        Mat normalize_(Mat src)
        {
            Cv2.CvtColor(src, src, ColorConversionCodes.BGR2RGB);

            Mat[] bgr = src.Split();
            for (int i = 0; i < bgr.Length; ++i)
            {
                bgr[i].ConvertTo(bgr[i], MatType.CV_32FC1, 1.0 / (255.0 * std[i]), (0.0 - mean[i]) / std[i]);
            }

            Cv2.Merge(bgr, src);

            foreach (Mat channel in bgr)
            {
                channel.Dispose();
            }

            return src;
        }

        unsafe void generate_image_feature(Mat srcimg)
        {
            Mat temp_image = new Mat();
            Cv2.Resize(srcimg, temp_image, new Size(inpWidth, inpHeight), 0, 0, InterpolationFlags.Cubic);
            Mat normalized_mat = normalize_(temp_image);
            Mat blob = CvDnn.BlobFromImage(normalized_mat);
            net.SetInput(blob);
            //模型推理，读取推理结果
            Mat[] outs = new Mat[1] { new Mat() };
            string[] outBlobNames = net.GetUnconnectedOutLayersNames().ToArray();
            net.Forward(outs, outBlobNames);
            float* ptr_feat = (float*)outs[0].Data;
            int len_image_feature = outs[0].Size(1);  //忽略第0维batchsize=1, len_image_feature是定值512,跟len_text_feature相等的, 也可以写死在类成员变量里
            image_features_input = new float[len_image_feature];
            float norm = 0.0f;
            for (int i = 0; i < len_image_feature; i++)
            {
                norm += ptr_feat[i] * ptr_feat[i];
            }
            norm = (float)Math.Sqrt(norm);
            for (int i = 0; i < len_image_feature; i++)
            {
                image_features_input[i] = ptr_feat[i] / norm;
            }
        }

        unsafe void generate_text_feature(List<string> texts)
        {
            List<List<int>> text_token = new List<List<int>>(texts.Count);
            for (int i = 0; i < texts.Count; i++)
            {
                text_token.Add(new List<int>());
            }

            for (int i = 0; i < texts.Count; i++)
            {
                tokenizer.encode_text(texts[i], text_token[i]);
            }

            if (text_token.Count * context_length > text_tokens_input.Length)
            {
                text_tokens_input = new int[text_token.Count * context_length];
            }

            foreach (int i in text_tokens_input) { text_tokens_input[i] = 0; }

            for (int i = 0; i < text_token.Count; i++)
            {
                if (text_token[i].Count > context_length)
                {
                    Console.WriteLine("text_features index " + i + " ,bigger than " + context_length + "\n");
                    continue;
                }
                for (int j = 0; j < text_token[i].Count; j++)
                {
                    text_tokens_input[i * context_length + j] = text_token[i][j];
                }

            }

            int[] text_token_shape = new int[] { 1, context_length };

            text_features_input = new float[text_token.Count, len_text_feature];

            long[] text_tokens_input_64 = new long[texts.Count * context_length];
            for (int i = 0; i < text_tokens_input_64.Length; i++)
            {
                text_tokens_input_64[i] = text_tokens_input[i];
            }

            for (int i = 0; i < text_token.Count; i++)
            {
                input_tensor = new DenseTensor<long>(text_tokens_input_64, new[] { 1, 52 });
                input_container.Clear();
                input_container.Add(NamedOnnxValue.CreateFromTensor("text", input_tensor));

                //运行 Inference 并获取结果
                result_infer = onnx_session.Run(input_container);

                // 将输出结果转为DisposableNamedOnnxValue数组
                results_onnxvalue = result_infer.ToArray();

                // 读取第一个节点输出并转为Tensor数据
                result_tensors = results_onnxvalue[0].AsTensor<float>();

                float[] text_feature_ptr = results_onnxvalue[0].AsTensor<float>().ToArray();

                float norm = 0.0f;
                for (int j = 0; j < len_text_feature; j++)
                {
                    norm += text_feature_ptr[j] * text_feature_ptr[j];
                }
                norm = (float)Math.Sqrt(norm);
                for (int j = 0; j < len_text_feature; j++)
                {
                    text_features_input[i, j] = text_feature_ptr[j] / norm;
                }

            }
        }

        void softmax(float[] input)
        {
            int length = input.Length;
            float[] exp_x = new float[length];
            float maxVal = input.Max();
            float sum = 0;
            for (int i = 0; i < length; i++)
            {
                float expval = (float)Math.Exp(input[i] - maxVal);
                exp_x[i] = expval;
                sum += expval;
            }
            for (int i = 0; i < length; i++)
            {
                input[i] = exp_x[i] / sum;
            }
        }

        int[] argsort_ascend(float[] array)
        {
            int array_len = array.Length;
            int[] array_index = new int[array_len];
            for (int i = 0; i < array_len; ++i)
            {
                array_index[i] = i;
            }

            var temp = array_index.ToList();

            temp.Sort((pos1, pos2) =>
             {

                 if (array[pos1] < array[pos2])
                 {
                     return -1;
                 }
                 else if (array[pos1] == array[pos2])
                 {
                     return 0;
                 }
                 else
                 {
                     return 0;
                 }

             });

            return temp.ToArray();
        }

        public List<Dictionary<string, float>> input_text_search_image(string text, float[] image_features, List<string> imglist)
        {

            int imgnum = imglist.Count;
            List<string> texts = new List<string> { text };

            generate_text_feature(texts);

            float[] logits_per_image = new float[imgnum];

            for (int i = 0; i < imgnum; i++)
            {
                float sum = 0;
                for (int j = 0; j < len_text_feature; j++)
                {
                    sum += image_features[i * len_text_feature + j] * text_features_input[0, j]; //图片特征向量跟文本特征向量做内积
                }
                logits_per_image[i] = 100 * sum;
            }

            softmax(logits_per_image);

            int[] index = argsort_ascend(logits_per_image);

            List<Dictionary<string, float>> top5imglist = new List<Dictionary<string, float>>(5);

            for (int i = 0; i < 5; i++)
            {
                int ind = index[imgnum - 1 - i];
                Dictionary<string, float> result = new Dictionary<string, float>();
                result.Add(imglist[ind], logits_per_image[ind]);
                top5imglist.Add(result);
            }
            return top5imglist;
        }

        public float[] generate_imagedir_features(string image_dir)
        {

            imglist = Common.listdir(image_dir);
            imgnum = imglist.Count;
            Console.WriteLine("遍历到" + imgnum + "张图片");

            float[] imagedir_features = new float[0];

            for (int i = 0; i < imgnum; i++)
            {
                string imgpath = imglist[i];

                Mat srcimg = Cv2.ImRead(imgpath);

                generate_image_feature(srcimg);

                imagedir_features = imagedir_features.Concat(image_features_input).ToArray();

                srcimg.Dispose();
            }

            return imagedir_features;

        }

    }

  1. public class Clip
  2.     {
  3.         int inpWidth = 224;
  4.         int inpHeight = 224;
  5.         float[] mean = new float[] { 0.48145466f, 0.4578275f, 0.40821073f };
  6.         float[] std = new float[] { 0.26862954f, 0.26130258f, 0.27577711f };
  7.  
  8.         int context_length = 52;
  9.         int len_text_feature = 512;
  10.  
  11.         Net net;
  12.         float[] image_features_input;
  13.  
  14.         SessionOptions options;
  15.         InferenceSession onnx_session;
  16.         Tensor<long> input_tensor;
  17.         List<NamedOnnxValue> input_container;
  18.         IDisposableReadOnlyCollection<DisposableNamedOnnxValue> result_infer;
  19.         DisposableNamedOnnxValue[] results_onnxvalue;
  20.         Tensor<float> result_tensors;
  21.  
  22.         TokenizerBase tokenizer;
  23.  
  24.         int[] text_tokens_input;
  25.         float[,] text_features_input;
  26.  
  27.         public int imgnum = 0;
  28.         public List<string> imglist = new List<string>();
  29.  
  30.         public Clip(string image_modelpath, string text_modelpath, string vocab_path)
  31.         {
  32.             net = CvDnn.ReadNetFromOnnx(image_modelpath);
  33.  
  34.             // 创建输出会话，用于输出模型读取信息
  35.             options = new SessionOptions();
  36.             options.LogSeverityLevel = OrtLoggingLevel.ORT_LOGGING_LEVEL_INFO;
  37.             options.AppendExecutionProvider_CPU(0);// 设置为CPU上运行
  38.             // 创建推理模型类，读取本地模型文件
  39.             onnx_session = new InferenceSession(text_modelpath, options);//model_path 为onnx模型文件的路径
  40.             // 创建输入容器
  41.             input_container = new List<NamedOnnxValue>();
  42.  
  43.             load_tokenizer(vocab_path);
  44.  
  45.         }
  46.  
  47.         void load_tokenizer(string vocab_path)
  48.         {
  49.  
  50.             tokenizer = new TokenizerClipChinese();
  51.             tokenizer.load_tokenize(vocab_path);
  52.             text_tokens_input = new int[1024 * context_length];
  53.         }
  54.  
  55.         Mat normalize_(Mat src)
  56.         {
  57.             Cv2.CvtColor(src, src, ColorConversionCodes.BGR2RGB);
  58.  
  59.             Mat[] bgr = src.Split();
  60.             for (int i = 0; i < bgr.Length; ++i)
  61.             {
  62.                 bgr[i].ConvertTo(bgr[i], MatType.CV_32FC1, 1.0 / (255.0 * std[i]), (0.0 - mean[i]) / std[i]);
  63.             }
  64.  
  65.             Cv2.Merge(bgr, src);
  66.  
  67.             foreach (Mat channel in bgr)
  68.             {
  69.                 channel.Dispose();
  70.             }
  71.  
  72.             return src;
  73.         }
  74.  
  75.         unsafe void generate_image_feature(Mat srcimg)
  76.         {
  77.             Mat temp_image = new Mat();
  78.             Cv2.Resize(srcimg, temp_image, new Size(inpWidth, inpHeight), 0, 0, InterpolationFlags.Cubic);
  79.             Mat normalized_mat = normalize_(temp_image);
  80.             Mat blob = CvDnn.BlobFromImage(normalized_mat);
  81.             net.SetInput(blob);
  82.             //模型推理，读取推理结果
  83.             Mat[] outs = new Mat[1] { new Mat() };
  84.             string[] outBlobNames = net.GetUnconnectedOutLayersNames().ToArray();
  85.             net.Forward(outs, outBlobNames);
  86.             float* ptr_feat = (float*)outs[0].Data;
  87.             int len_image_feature = outs[0].Size(1);  //忽略第0维batchsize=1, len_image_feature是定值512,跟len_text_feature相等的, 也可以写死在类成员变量里
  88.             image_features_input = new float[len_image_feature];
  89.             float norm = 0.0f;
  90.             for (int i = 0; i < len_image_feature; i++)
  91.             {
  92.                 norm += ptr_feat[i] * ptr_feat[i];
  93.             }
  94.             norm = (float)Math.Sqrt(norm);
  95.             for (int i = 0; i < len_image_feature; i++)
  96.             {
  97.                 image_features_input[i] = ptr_feat[i] / norm;
  98.             }
  99.         }
  100.  
  101.         unsafe void generate_text_feature(List<string> texts)
  102.         {
  103.             List<List<int>> text_token = new List<List<int>>(texts.Count);
  104.             for (int i = 0; i < texts.Count; i++)
  105.             {
  106.                 text_token.Add(new List<int>());
  107.             }
  108.  
  109.             for (int i = 0; i < texts.Count; i++)
  110.             {
  111.                 tokenizer.encode_text(texts[i], text_token[i]);
  112.             }
  113.  
  114.             if (text_token.Count * context_length > text_tokens_input.Length)
  115.             {
  116.                 text_tokens_input = new int[text_token.Count * context_length];
  117.             }
  118.  
  119.             foreach (int i in text_tokens_input) { text_tokens_input[i] = 0; }
  120.  
  121.             for (int i = 0; i < text_token.Count; i++)
  122.             {
  123.                 if (text_token[i].Count > context_length)
  124.                 {
  125.                     Console.WriteLine("text_features index " + i + " ,bigger than " + context_length + "\n");
  126.                     continue;
  127.                 }
  128.                 for (int j = 0; j < text_token[i].Count; j++)
  129.                 {
  130.                     text_tokens_input[i * context_length + j] = text_token[i][j];
  131.                 }
  132.  
  133.             }
  134.  
  135.             int[] text_token_shape = new int[] { 1, context_length };
  136.  
  137.             text_features_input = new float[text_token.Count, len_text_feature];
  138.  
  139.             long[] text_tokens_input_64 = new long[texts.Count * context_length];
  140.             for (int i = 0; i < text_tokens_input_64.Length; i++)
  141.             {
  142.                 text_tokens_input_64[i] = text_tokens_input[i];
  143.             }
  144.  
  145.             for (int i = 0; i < text_token.Count; i++)
  146.             {
  147.                 input_tensor = new DenseTensor<long>(text_tokens_input_64, new[] { 1, 52 });
  148.                 input_container.Clear();
  149.                 input_container.Add(NamedOnnxValue.CreateFromTensor("text", input_tensor));
  150.  
  151.                 //运行 Inference 并获取结果
  152.                 result_infer = onnx_session.Run(input_container);
  153.  
  154.                 // 将输出结果转为DisposableNamedOnnxValue数组
  155.                 results_onnxvalue = result_infer.ToArray();
  156.  
  157.                 // 读取第一个节点输出并转为Tensor数据
  158.                 result_tensors = results_onnxvalue[0].AsTensor<float>();
  159.  
  160.                 float[] text_feature_ptr = results_onnxvalue[0].AsTensor<float>().ToArray();
  161.  
  162.                 float norm = 0.0f;
  163.                 for (int j = 0; j < len_text_feature; j++)
  164.                 {
  165.                     norm += text_feature_ptr[j] * text_feature_ptr[j];
  166.                 }
  167.                 norm = (float)Math.Sqrt(norm);
  168.                 for (int j = 0; j < len_text_feature; j++)
  169.                 {
  170.                     text_features_input[i, j] = text_feature_ptr[j] / norm;
  171.                 }
  172.  
  173.             }
  174.         }
  175.  
  176.         void softmax(float[] input)
  177.         {
  178.             int length = input.Length;
  179.             float[] exp_x = new float[length];
  180.             float maxVal = input.Max();
  181.             float sum = 0;
  182.             for (int i = 0; i < length; i++)
  183.             {
  184.                 float expval = (float)Math.Exp(input[i] - maxVal);
  185.                 exp_x[i] = expval;
  186.                 sum += expval;
  187.             }
  188.             for (int i = 0; i < length; i++)
  189.             {
  190.                 input[i] = exp_x[i] / sum;
  191.             }
  192.         }
  193.  
  194.         int[] argsort_ascend(float[] array)
  195.         {
  196.             int array_len = array.Length;
  197.             int[] array_index = new int[array_len];
  198.             for (int i = 0; i < array_len; ++i)
  199.             {
  200.                 array_index[i] = i;
  201.             }
  202.  
  203.             var temp = array_index.ToList();
  204.  
  205.             temp.Sort((pos1, pos2) =>
  206.              {
  207.  
  208.                  if (array[pos1] < array[pos2])
  209.                  {
  210.                      return -1;
  211.                  }
  212.                  else if (array[pos1] == array[pos2])
  213.                  {
  214.                      return 0;
  215.                  }
  216.                  else
  217.                  {
  218.                      return 0;
  219.                  }
  220.  
  221.              });
  222.  
  223.             return temp.ToArray();
  224.         }
  225.  
  226.         public List<Dictionary<string, float>> input_text_search_image(string text, float[] image_features, List<string> imglist)
  227.         {
  228.  
  229.             int imgnum = imglist.Count;
  230.             List<string> texts = new List<string> { text };
  231.  
  232.             generate_text_feature(texts);
  233.  
  234.             float[] logits_per_image = new float[imgnum];
  235.  
  236.             for (int i = 0; i < imgnum; i++)
  237.             {
  238.                 float sum = 0;
  239.                 for (int j = 0; j < len_text_feature; j++)
  240.                 {
  241.                     sum += image_features[i * len_text_feature + j] * text_features_input[0, j]; //图片特征向量跟文本特征向量做内积
  242.                 }
  243.                 logits_per_image[i] = 100 * sum;
  244.             }
  245.  
  246.             softmax(logits_per_image);
  247.  
  248.             int[] index = argsort_ascend(logits_per_image);
  249.  
  250.             List<Dictionary<string, float>> top5imglist = new List<Dictionary<string, float>>(5);
  251.  
  252.             for (int i = 0; i < 5; i++)
  253.             {
  254.                 int ind = index[imgnum - 1 - i];
  255.                 Dictionary<string, float> result = new Dictionary<string, float>();
  256.                 result.Add(imglist[ind], logits_per_image[ind]);
  257.                 top5imglist.Add(result);
  258.             }
  259.             return top5imglist;
  260.         }
  261.  
  262.         public float[] generate_imagedir_features(string image_dir)
  263.         {
  264.  
  265.             imglist = Common.listdir(image_dir);
  266.             imgnum = imglist.Count;
  267.             Console.WriteLine("遍历到" + imgnum + "张图片");
  268.  
  269.             float[] imagedir_features = new float[0];
  270.  
  271.             for (int i = 0; i < imgnum; i++)
  272.             {
  273.                 string imgpath = imglist[i];
  274.  
  275.                 Mat srcimg = Cv2.ImRead(imgpath);
  276.  
  277.                 generate_image_feature(srcimg);
  278.  
  279.                 imagedir_features = imagedir_features.Concat(image_features_input).ToArray();
  280.  
  281.                 srcimg.Dispose();
  282.             }
  283.  
  284.             return imagedir_features;
  285.  
  286.         }
  287.  
  288.     }

下载

源码下载

声明：本文内容由网友自发贡献，不代表【wpsshop博客】立场，版权归原作者所有，本站不承担相应法律责任。如您发现有侵权的内容，请联系我们。转载请注明出处：https://www.wpsshop.cn/w/代码探险家/article/detail/775624
推荐阅读
相关标签

Copyright © 2003-2013 www.wpsshop.cn 版权所有，并保留所有权利。

  

闽ICP备14008679号