技能备份 - 2026-04-15 (40个技能)
This commit is contained in:
root
@@ -0,0 +1,131 @@
|
||||
---
|
||||
name: tencentcloud-ocr-general
|
||||
description: 腾讯云广告文字识别(AdvertiseOCR)接口调用技能。当用户需要从图片中识别文字内容时,应使用此技能。支持中英文、横排、竖排及倾斜场景的图片文字识别,支持90度、180度、270度翻转场景的图片识别,返回文本框位置与文字内容。支持图片Base64和URL两种输入方式。
|
||||
---
|
||||
|
||||
# 腾讯云广告文字识别 (AdvertiseOCR)
|
||||
|
||||
## 用途
|
||||
|
||||
调用腾讯云OCR广告文字识别接口，支持图片内文字的检测和识别，返回文本框位置与文字内容。具有较高召回率和准确率。
|
||||
|
||||
核心能力：
|
||||
- **中英文识别**：支持中英文混合文字识别
|
||||
- **多方向支持**：支持横排、竖排以及倾斜场景文字识别
|
||||
- **翻转支持**：支持90度、180度、270度翻转场景文字识别
|
||||
- **坐标返回**：返回每个文本行的四顶点坐标（Polygon）
|
||||
- **置信度评估**：返回每个文本行的识别置信度（0~100）
|
||||
|
||||
官方文档：https://cloud.tencent.com/document/api/866/49524
|
||||
|
||||
默认接口请求频率限制：20次/秒。
|
||||
|
||||
## 使用时机
|
||||
|
||||
当用户提出以下需求时触发此技能：
|
||||
- 需要从图片中提取文字信息
|
||||
- 需要识别图片上的文字内容
|
||||
- 涉及文字OCR识别的任何场景
|
||||
- 需要获取图片中文字的位置坐标信息
|
||||
|
||||
## 环境要求
|
||||
|
||||
- Python 3.6+
|
||||
- 依赖：`tencentcloud-sdk-python`（通过 `pip install tencentcloud-sdk-python` 安装）
|
||||
- 环境变量：
|
||||
- `TENCENTCLOUD_SECRET_ID`：腾讯云API密钥ID
|
||||
- `TENCENTCLOUD_SECRET_KEY`：腾讯云API密钥Key
|
||||
|
||||
## 使用方式
|
||||
|
||||
运行 `scripts/main.py` 脚本完成图片文字识别。
|
||||
|
||||
### 请求参数
|
||||
|
||||
| 参数 | 类型 | 必填 | 说明 |
|
||||
|------|------|------|------|
|
||||
| ImageBase64 | str | 否(二选一) | 图片Base64值，编码后不超过10M，分辨率建议600*800以上，支持PNG/JPG/JPEG/BMP |
|
||||
| ImageUrl | str | 否(二选一) | 图片URL地址，建议存储于腾讯云COS。都提供时仅使用ImageUrl |
|
||||
|
||||
### 输出格式
|
||||
|
||||
识别成功后返回 JSON 格式结果：
|
||||
|
||||
```json
|
||||
{
|
||||
"TextDetections": [
|
||||
{
|
||||
"DetectedText": "识别出的文本行内容",
|
||||
"Confidence": 99,
|
||||
"Polygon": [
|
||||
{"X": 0, "Y": 0},
|
||||
{"X": 100, "Y": 0},
|
||||
{"X": 100, "Y": 50},
|
||||
{"X": 0, "Y": 50}
|
||||
],
|
||||
"AdvancedInfo": "{\"Parag\":{\"ParagNo\":1}}"
|
||||
}
|
||||
],
|
||||
"TextCount": 1,
|
||||
"ImageSize": {
|
||||
"Width": 800,
|
||||
"Height": 600
|
||||
},
|
||||
"RequestId": "xxx"
|
||||
}
|
||||
```
|
||||
|
||||
**响应字段说明：**
|
||||
|
||||
| 字段 | 类型 | 说明 |
|
||||
|------|------|------|
|
||||
| TextDetections | list | 检测到的文本信息列表 |
|
||||
| TextDetections[].DetectedText | str | 识别出的文本行内容 |
|
||||
| TextDetections[].Confidence | int | 置信度 0~100 |
|
||||
| TextDetections[].Polygon | list of Coord | 文本行坐标，四个顶点坐标（X, Y） |
|
||||
| TextDetections[].AdvancedInfo | str | 扩展字段，含段落信息Parag(ParagNo) |
|
||||
| TextCount | int | 检测到的文本行数量 |
|
||||
| ImageSize | object | 图片分辨率信息，含Width和Height（单位px） |
|
||||
| RequestId | str | 唯一请求ID |
|
||||
|
||||
### 错误码说明
|
||||
|
||||
| 错误码 | 含义 |
|
||||
|--------|------|
|
||||
| FailedOperation.DownLoadError | 文件下载失败 |
|
||||
| FailedOperation.EmptyImageError | 图片内容为空 |
|
||||
| FailedOperation.EngineRecognizeTimeout | 引擎识别超时 |
|
||||
| FailedOperation.ImageDecodeFailed | 图片解码失败 |
|
||||
| FailedOperation.ImageNoText | 图片中未检测到文本 |
|
||||
| FailedOperation.LanguageNotSupport | 输入的Language不支持 |
|
||||
| FailedOperation.OcrFailed | OCR识别失败 |
|
||||
| FailedOperation.UnKnowError | 未知错误 |
|
||||
| FailedOperation.UnOpenError | 服务未开通 |
|
||||
| InvalidParameterValue.InvalidParameterValueLimit | 参数值错误 |
|
||||
| LimitExceeded.TooLargeFileError | 文件内容太大 |
|
||||
| ResourceUnavailable.InArrears | 账号已欠费 |
|
||||
| ResourceUnavailable.ResourcePackageRunOut | 账号资源包耗尽 |
|
||||
| ResourcesSoldOut.ChargeStatusException | 计费状态异常 |
|
||||
|
||||
### 业务逻辑说明
|
||||
|
||||
1. ImageBase64和ImageUrl必须提供其一，都提供时只使用ImageUrl
|
||||
2. 图片经Base64编码后不超过10M，分辨率建议600*800以上
|
||||
3. 支持PNG、JPG、JPEG、BMP格式
|
||||
4. 始终计费
|
||||
|
||||
### 调用示例
|
||||
|
||||
```bash
|
||||
# 通过URL识别图片文字
|
||||
python scripts/main.py --image-url "https://example.com/ad_image.jpg"
|
||||
|
||||
# 通过文件路径(自动Base64编码)识别
|
||||
python scripts/main.py --image-base64 ./ad_image.jpg
|
||||
|
||||
# 通过Base64文本文件识别
|
||||
python scripts/main.py --image-base64 ./base64.txt
|
||||
|
||||
# 指定地域
|
||||
python scripts/main.py --image-url "https://example.com/ad_image.jpg" --region ap-beijing
|
||||
```
|
||||
@@ -0,0 +1,6 @@
|
||||
{
|
||||
"ownerId": "kn71qd8mtdsrf1bsagg94kx5tn82dqt6",
|
||||
"slug": "tencentcloud-ocr-general",
|
||||
"version": "1.0.1",
|
||||
"publishedAt": 1772961532647
|
||||
}
|
||||
@@ -0,0 +1,229 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
腾讯云广告文字识别(AdvertiseOCR)调用脚本
|
||||
|
||||
支持广告商品图片内文字的检测和识别，返回文本框位置与文字内容。
|
||||
支持中英文、横排、竖排以及倾斜场景文字识别，支持90度、180度、270度翻转。
|
||||
需要环境变量: TENCENTCLOUD_SECRET_ID, TENCENTCLOUD_SECRET_KEY
|
||||
|
||||
用法:
|
||||
python main.py --image-url <url>
|
||||
python main.py --image-base64 <base64_or_filepath>
|
||||
"""
|
||||
|
||||
import argparse
|
||||
import json
|
||||
import os
|
||||
import sys
|
||||
import base64
|
||||
|
||||
# SDK 最大图片限制 (10MB)
|
||||
MAX_IMAGE_SIZE_BYTES = 10 * 1024 * 1024
|
||||
|
||||
# 错误码含义映射
|
||||
ERROR_CODE_MAP = {
|
||||
"FailedOperation.DownLoadError": "文件下载失败",
|
||||
"FailedOperation.EmptyImageError": "图片内容为空",
|
||||
"FailedOperation.EngineRecognizeTimeout": "引擎识别超时",
|
||||
"FailedOperation.ImageDecodeFailed": "图片解码失败",
|
||||
"FailedOperation.ImageNoText": "图片中未检测到文本",
|
||||
"FailedOperation.LanguageNotSupport": "输入的Language不支持",
|
||||
"FailedOperation.OcrFailed": "OCR识别失败",
|
||||
"FailedOperation.UnKnowError": "未知错误",
|
||||
"FailedOperation.UnOpenError": "服务未开通",
|
||||
"InvalidParameterValue.InvalidParameterValueLimit": "参数值错误",
|
||||
"LimitExceeded.TooLargeFileError": "文件内容太大",
|
||||
"ResourceUnavailable.InArrears": "账号已欠费",
|
||||
"ResourceUnavailable.ResourcePackageRunOut": "账号资源包耗尽",
|
||||
"ResourcesSoldOut.ChargeStatusException": "计费状态异常",
|
||||
}
|
||||
|
||||
|
||||
def validate_env() -> tuple:
|
||||
"""校验并返回腾讯云API密钥。"""
|
||||
secret_id = os.environ.get("TENCENTCLOUD_SECRET_ID")
|
||||
secret_key = os.environ.get("TENCENTCLOUD_SECRET_KEY")
|
||||
if not secret_id or not secret_key:
|
||||
print("错误: 请设置环境变量 TENCENTCLOUD_SECRET_ID 和 TENCENTCLOUD_SECRET_KEY", file=sys.stderr)
|
||||
sys.exit(1)
|
||||
return secret_id, secret_key
|
||||
|
||||
|
||||
def load_image_base64(value: str) -> str:
|
||||
"""
|
||||
加载 Base64 图片内容。
|
||||
如果 value 是一个存在的文件路径，则读取文件内容作为 Base64；
|
||||
否则直接视为 Base64 字符串。
|
||||
"""
|
||||
if os.path.isfile(value):
|
||||
with open(value, "rb") as f:
|
||||
raw = f.read()
|
||||
# 如果文件内容本身就是Base64文本(如txt文件)，直接使用
|
||||
try:
|
||||
raw_str = raw.decode("utf-8").strip()
|
||||
base64.b64decode(raw_str, validate=True)
|
||||
return raw_str
|
||||
except Exception:
|
||||
pass
|
||||
# 否则将二进制文件编码为Base64
|
||||
if len(raw) > MAX_IMAGE_SIZE_BYTES:
|
||||
print(f"错误: 图片文件大小超过 {MAX_IMAGE_SIZE_BYTES // (1024 * 1024)}MB 限制", file=sys.stderr)
|
||||
sys.exit(1)
|
||||
encoded = base64.b64encode(raw).decode("utf-8")
|
||||
return encoded
|
||||
else:
|
||||
# 直接作为 Base64 字符串使用
|
||||
try:
|
||||
decoded = base64.b64decode(value, validate=True)
|
||||
if len(decoded) > MAX_IMAGE_SIZE_BYTES:
|
||||
print(f"错误: 图片大小超过 {MAX_IMAGE_SIZE_BYTES // (1024 * 1024)}MB 限制", file=sys.stderr)
|
||||
sys.exit(1)
|
||||
except Exception:
|
||||
print("错误: 提供的 ImageBase64 不是合法的 Base64 编码，也不是有效的文件路径", file=sys.stderr)
|
||||
sys.exit(1)
|
||||
return value
|
||||
|
||||
|
||||
def format_response(resp_json: dict) -> dict:
|
||||
"""格式化响应结果，提取关键信息并结构化输出。"""
|
||||
output = {}
|
||||
|
||||
# 文本检测结果
|
||||
text_detections = resp_json.get("TextDetections")
|
||||
if text_detections:
|
||||
formatted_texts = []
|
||||
for item in text_detections:
|
||||
text_info = {
|
||||
"DetectedText": item.get("DetectedText", ""),
|
||||
"Confidence": item.get("Confidence", 0),
|
||||
}
|
||||
# 文本行坐标
|
||||
polygon = item.get("Polygon")
|
||||
if polygon:
|
||||
text_info["Polygon"] = polygon
|
||||
# 扩展信息
|
||||
advanced_info = item.get("AdvancedInfo")
|
||||
if advanced_info:
|
||||
text_info["AdvancedInfo"] = advanced_info
|
||||
formatted_texts.append(text_info)
|
||||
output["TextDetections"] = formatted_texts
|
||||
output["TextCount"] = len(formatted_texts)
|
||||
|
||||
# 图片分辨率
|
||||
image_size = resp_json.get("ImageSize")
|
||||
if image_size:
|
||||
output["ImageSize"] = image_size
|
||||
|
||||
# 请求ID
|
||||
if "RequestId" in resp_json:
|
||||
output["RequestId"] = resp_json["RequestId"]
|
||||
|
||||
return output
|
||||
|
||||
|
||||
def call_advertise_ocr(args: argparse.Namespace) -> None:
|
||||
"""调用腾讯云 AdvertiseOCR 接口。"""
|
||||
try:
|
||||
from tencentcloud.common import credential
|
||||
from tencentcloud.common.profile.client_profile import ClientProfile
|
||||
from tencentcloud.common.profile.http_profile import HttpProfile
|
||||
from tencentcloud.common.exception.tencent_cloud_sdk_exception import TencentCloudSDKException
|
||||
from tencentcloud.ocr.v20181119 import ocr_client, models
|
||||
except ImportError:
|
||||
print("错误: 缺少依赖 tencentcloud-sdk-python，请执行: pip install tencentcloud-sdk-python", file=sys.stderr)
|
||||
sys.exit(1)
|
||||
|
||||
secret_id, secret_key = validate_env()
|
||||
|
||||
# 构建客户端
|
||||
cred = credential.Credential(secret_id, secret_key)
|
||||
http_profile = HttpProfile()
|
||||
http_profile.endpoint = "ocr.tencentcloudapi.com"
|
||||
client_profile = ClientProfile()
|
||||
client_profile.httpProfile = http_profile
|
||||
region = args.region if args.region else "ap-guangzhou"
|
||||
client = ocr_client.OcrClient(cred, region, client_profile)
|
||||
|
||||
# 构建请求
|
||||
req = models.AdvertiseOCRRequest()
|
||||
|
||||
if args.image_url:
|
||||
req.ImageUrl = args.image_url
|
||||
elif args.image_base64:
|
||||
req.ImageBase64 = load_image_base64(args.image_base64)
|
||||
else:
|
||||
print("错误: 必须提供 --image-url 或 --image-base64 之一", file=sys.stderr)
|
||||
sys.exit(1)
|
||||
|
||||
# 发起请求
|
||||
try:
|
||||
resp = client.AdvertiseOCR(req)
|
||||
except TencentCloudSDKException as e:
|
||||
error_desc = ERROR_CODE_MAP.get(e.code, "")
|
||||
error_msg = f"API调用失败 [{e.code}]: {e.message}"
|
||||
if error_desc:
|
||||
error_msg += f" ({error_desc})"
|
||||
print(error_msg, file=sys.stderr)
|
||||
if e.requestId:
|
||||
print(f"RequestId: {e.requestId}", file=sys.stderr)
|
||||
sys.exit(1)
|
||||
|
||||
# 解析并格式化输出
|
||||
resp_json = json.loads(resp.to_json_string())
|
||||
result = format_response(resp_json)
|
||||
print(json.dumps(result, ensure_ascii=False, indent=2))
|
||||
|
||||
|
||||
def build_parser() -> argparse.ArgumentParser:
|
||||
"""构建命令行参数解析器。"""
|
||||
parser = argparse.ArgumentParser(
|
||||
description="腾讯云广告文字识别(AdvertiseOCR)调用工具",
|
||||
formatter_class=argparse.RawDescriptionHelpFormatter,
|
||||
epilog="""
|
||||
示例:
|
||||
# 通过URL识别广告图片文字
|
||||
python main.py --image-url "https://example.com/ad_image.jpg"
|
||||
|
||||
# 通过文件路径(自动Base64编码)识别
|
||||
python main.py --image-base64 ./ad_image.jpg
|
||||
|
||||
# 通过Base64文本文件识别
|
||||
python main.py --image-base64 ./base64.txt
|
||||
|
||||
# 指定地域
|
||||
python main.py --image-url "https://example.com/ad_image.jpg" --region ap-beijing
|
||||
""",
|
||||
)
|
||||
|
||||
# 图片输入（二选一）
|
||||
img_group = parser.add_mutually_exclusive_group(required=True)
|
||||
img_group.add_argument(
|
||||
"--image-url",
|
||||
type=str,
|
||||
help="图片URL地址，建议存储于腾讯云COS",
|
||||
)
|
||||
img_group.add_argument(
|
||||
"--image-base64",
|
||||
type=str,
|
||||
help="图片Base64字符串，或图片/Base64文本文件的路径",
|
||||
)
|
||||
|
||||
# 可选参数
|
||||
parser.add_argument(
|
||||
"--region",
|
||||
type=str,
|
||||
default=None,
|
||||
help="腾讯云地域，默认 ap-guangzhou",
|
||||
)
|
||||
|
||||
return parser
|
||||
|
||||
|
||||
def main():
|
||||
parser = build_parser()
|
||||
args = parser.parse_args()
|
||||
call_advertise_ocr(args)
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
Reference in New Issue
Block a user