Google表格：自动获取图像分类标签的公式，脚本或扩展名？

Question

我有一个带有图像URL列表的Google表格。

[第一列：图片网址。第二列：描述图像的标签（例如：“景观，山，田野，日落，湖泊”，用于包含山，田野，日落等的风景图像）

我想用“ ImageTags（ImageUrl）”之类的公式自动填充第2列。

是否有任何公式，扩展名或脚本（已经可用：））？

谢谢！

Answer 1

选项是使用Google Cloud Vision API标记图像。您可以测试API以确定它是否适合您的需求here。

有两个使用Google Apps脚本来调用API的教程，可以帮助您开发所需的确切代码：

1. Using a service account to authenticate, detect labels and send a GMail message

2. Using oAuth2 to authenticate and label images

作为示例，使用以下图像：

并使用以下参数调用API：

• HTTP请求

POST https://vision.googleapis.com/v1/images:annotate

• 正文

{
  "requests": [
    {
      "image": {
        "source": {
          "imageUri": "https://i.stack.imgur.com/4vwKt.jpg"
        }
      },
      "features": [
        {
          "type": "LABEL_DETECTION"
        }
      ]
    }
  ]
}

您获得以下结果（2019年11月5日）：

{
  "responses": [
    {
      "labelAnnotations": [
        {
          "mid": "/m/09j06",
          "description": "Hot air balloon",
          "score": 0.9889263,
          "topicality": 0.9889263
        },
        {
          "mid": "/m/01j51",
          "description": "Balloon",
          "score": 0.95322704,
          "topicality": 0.95322704
        },
        {
          "mid": "/m/02p81ht",
          "description": "Hot air ballooning",
          "score": 0.9223063,
          "topicality": 0.9223063
        },
        {
          "mid": "/m/0cmqr_4",
          "description": "Party supply",
          "score": 0.9016293,
          "topicality": 0.9016293
        },
        {
          "mid": "/m/01d40f",
          "description": "Dress",
          "score": 0.86037284,
          "topicality": 0.86037284
        },
        {
          "mid": "/m/07yv9",
          "description": "Vehicle",
          "score": 0.7725018,
          "topicality": 0.7725018
        },
        {
          "mid": "/m/01bqvp",
          "description": "Sky",
          "score": 0.7326111,
          "topicality": 0.7326111
        },
        {
          "mid": "/m/0ds99lh",
          "description": "Fun",
          "score": 0.7039424,
          "topicality": 0.7039424
        },
        {
          "mid": "/m/016pp7",
          "description": "Happy",
          "score": 0.6789371,
          "topicality": 0.6789371
        },
        {
          "mid": "/m/06bm2",
          "description": "Recreation",
          "score": 0.6695586,
          "topicality": 0.6695586
        }
      ]
    }
  ]
}