Compare commits
2 Commits
0c2bb58cca
...
572ba946fd
| Author | SHA1 | Date |
|---|---|---|
 ljw
|
572ba946fd | |
|
ljw
|
6f30570604 |
|
|
@ -125,6 +125,11 @@
|
|||
<artifactId>aspectjweaver</artifactId>
|
||||
<version>1.9.7</version>
|
||||
</dependency>
|
||||
<dependency>
|
||||
<groupId>com.volcengine</groupId>
|
||||
<artifactId>volcengine-java-sdk-ark-runtime</artifactId>
|
||||
<version>LATEST</version>
|
||||
</dependency>
|
||||
|
||||
</dependencies>
|
||||
|
||||
|
|
|
|||
|
|
@ -0,0 +1,520 @@
|
|||
# 豆包垃圾分析接口文档
|
||||
|
||||
## 接口概述
|
||||
|
||||
本接口基于火山引擎豆包AI模型，提供垃圾车辆图片分析功能。通过上传多张垃圾车辆图片，AI会自动识别车棚状态、垃圾类型、轻物质和重物质占比等信息。
|
||||
|
||||
**接口路径：** `/douBao/analyzeGarbage`
|
||||
**请求方式：** `POST`
|
||||
**Content-Type：** `application/json`
|
||||
|
||||
---
|
||||
|
||||
## 请求参数
|
||||
|
||||
### 请求体（GarbageAnalysisParam）
|
||||
|
||||
| 参数名 | 类型 | 必填 | 说明 |
|
||||
|--------|------|------|------|
|
||||
| imageUrls | List<String> | 是 | 图片URL列表，支持多张图片，至少需要一张 |
|
||||
| orderSn | String | 否 | 订单号，用于关联订单信息 |
|
||||
| remark | String | 否 | 备注信息，可提供额外的分析上下文 |
|
||||
|
||||
### 参数说明
|
||||
|
||||
- **imageUrls**：图片URL列表，支持HTTP/HTTPS协议的图片链接。建议使用可公开访问的图片URL。
|
||||
- **orderSn**：订单号，用于在分析结果中关联订单信息，便于后续查询和追踪。
|
||||
- **remark**：备注信息，可以包含订单相关的额外信息，如装货地点、卸货地点等，有助于AI更准确地进行分析。
|
||||
|
||||
---
|
||||
|
||||
## 响应结果
|
||||
|
||||
### 响应体（GarbageAnalysisResult）
|
||||
|
||||
| 字段名 | 类型 | 说明 |
|
||||
|--------|------|------|
|
||||
| isShedOpened | Boolean | 车棚是否完全打开。true表示已打开，false表示未打开 |
|
||||
| details | String | 明细信息，车厢内包含的物品列表。格式：物品1(数量描述),物品2(数量描述)，例如：石块(多),木材(少) |
|
||||
| lightMaterialPercentage | BigDecimal | 轻物质占比（百分比，0-100） |
|
||||
| heavyMaterialPercentage | BigDecimal | 重物质占比（百分比，0-100） |
|
||||
| garbageType | String | 垃圾类型，可能的值：拆除垃圾、装修垃圾、抛货 |
|
||||
| judgmentBasis | String | 判断依据，说明AI判断垃圾类型的原因 |
|
||||
| analysisTime | Long | 分析时间戳（毫秒） |
|
||||
| rawResponse | String | AI原始响应内容，用于调试和问题排查 |
|
||||
| garbageTypes | List<GarbageTypeInfo> | 垃圾类型分析列表（保留字段，用于兼容） |
|
||||
| totalWeight | BigDecimal | 总重量（单位：千克）（保留字段，用于兼容） |
|
||||
|
||||
### GarbageTypeInfo 对象结构（保留字段）
|
||||
|
||||
| 字段名 | 类型 | 说明 |
|
||||
|--------|------|------|
|
||||
| typeName | String | 垃圾类型名称 |
|
||||
| category | String | 具体垃圾种类 |
|
||||
| judgmentBasis | String | 判断依据 |
|
||||
| percentage | BigDecimal | 占比（百分比，0-100） |
|
||||
| weight | BigDecimal | 估算重量（单位：千克） |
|
||||
|
||||
---
|
||||
|
||||
## 请求示例
|
||||
|
||||
### 场景一：最小参数（仅图片URL）
|
||||
|
||||
**适用场景：** 快速测试，仅提供必要的图片URL
|
||||
|
||||
```json
|
||||
{
|
||||
"imageUrls": [
|
||||
"https://example.com/garbage-truck-1.jpg"
|
||||
]
|
||||
}
|
||||
```
|
||||
|
||||
**cURL 命令：**
|
||||
```bash
|
||||
curl -X POST "http://localhost:8080/douBao/analyzeGarbage" \
|
||||
-H "Content-Type: application/json" \
|
||||
-d '{
|
||||
"imageUrls": [
|
||||
"https://example.com/garbage-truck-1.jpg"
|
||||
]
|
||||
}'
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### 场景二：单张图片 + 订单号
|
||||
|
||||
**适用场景：** 关联订单信息
|
||||
|
||||
```json
|
||||
{
|
||||
"imageUrls": [
|
||||
"https://example.com/garbage-truck-1.jpg"
|
||||
],
|
||||
"orderSn": "ORD20240101001"
|
||||
}
|
||||
```
|
||||
|
||||
**cURL 命令：**
|
||||
```bash
|
||||
curl -X POST "http://localhost:8080/douBao/analyzeGarbage" \
|
||||
-H "Content-Type: application/json" \
|
||||
-d '{
|
||||
"imageUrls": [
|
||||
"https://example.com/garbage-truck-1.jpg"
|
||||
],
|
||||
"orderSn": "ORD20240101001"
|
||||
}'
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### 场景三：多张图片（推荐）
|
||||
|
||||
**适用场景：** 从不同角度拍摄，提高分析准确性
|
||||
|
||||
```json
|
||||
{
|
||||
"imageUrls": [
|
||||
"https://example.com/garbage-truck-front.jpg",
|
||||
"https://example.com/garbage-truck-side.jpg",
|
||||
"https://example.com/garbage-truck-back.jpg"
|
||||
],
|
||||
"orderSn": "ORD20240101001",
|
||||
"remark": "装货地点：南京市建邺区，卸货地点：南京市江宁区"
|
||||
}
|
||||
```
|
||||
|
||||
**cURL 命令：**
|
||||
```bash
|
||||
curl -X POST "http://localhost:8080/douBao/analyzeGarbage" \
|
||||
-H "Content-Type: application/json" \
|
||||
-d '{
|
||||
"imageUrls": [
|
||||
"https://example.com/garbage-truck-front.jpg",
|
||||
"https://example.com/garbage-truck-side.jpg",
|
||||
"https://example.com/garbage-truck-back.jpg"
|
||||
],
|
||||
"orderSn": "ORD20240101001",
|
||||
"remark": "装货地点：南京市建邺区，卸货地点：南京市江宁区"
|
||||
}'
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### 场景四：完整参数示例
|
||||
|
||||
**适用场景：** 生产环境，包含所有可选信息
|
||||
|
||||
```json
|
||||
{
|
||||
"imageUrls": [
|
||||
"https://cdn.example.com/orders/2024/01/01/truck-front-001.jpg",
|
||||
"https://cdn.example.com/orders/2024/01/01/truck-side-001.jpg"
|
||||
],
|
||||
"orderSn": "ORD20240101001",
|
||||
"remark": "订单编号：ORD20240101001；装货地点：南京市建邺区奥体中心；卸货地点：南京市江宁区垃圾处理厂；装货时间：2024-01-01 10:00:00"
|
||||
}
|
||||
```
|
||||
|
||||
**cURL 命令：**
|
||||
```bash
|
||||
curl -X POST "http://localhost:8080/douBao/analyzeGarbage" \
|
||||
-H "Content-Type: application/json" \
|
||||
-d '{
|
||||
"imageUrls": [
|
||||
"https://cdn.example.com/orders/2024/01/01/truck-front-001.jpg",
|
||||
"https://cdn.example.com/orders/2024/01/01/truck-side-001.jpg"
|
||||
],
|
||||
"orderSn": "ORD20240101001",
|
||||
"remark": "订单编号：ORD20240101001；装货地点：南京市建邺区奥体中心；卸货地点：南京市江宁区垃圾处理厂；装货时间：2024-01-01 10:00:00"
|
||||
}'
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### Java 示例
|
||||
|
||||
#### 方式一：使用链式调用（推荐）
|
||||
|
||||
```java
|
||||
import com.njzscloud.supervisory.doubao.param.GarbageAnalysisParam;
|
||||
import com.njzscloud.supervisory.doubao.result.GarbageAnalysisResult;
|
||||
import com.njzscloud.common.core.utils.R;
|
||||
import java.util.Arrays;
|
||||
|
||||
// 最小参数
|
||||
GarbageAnalysisParam param1 = new GarbageAnalysisParam()
|
||||
.setImageUrls(Arrays.asList("https://example.com/garbage-truck-1.jpg"));
|
||||
|
||||
// 完整参数
|
||||
GarbageAnalysisParam param2 = new GarbageAnalysisParam()
|
||||
.setImageUrls(Arrays.asList(
|
||||
"https://example.com/garbage-truck-front.jpg",
|
||||
"https://example.com/garbage-truck-side.jpg"
|
||||
))
|
||||
.setOrderSn("ORD20240101001")
|
||||
.setRemark("装货地点：南京市建邺区，卸货地点：南京市江宁区");
|
||||
|
||||
R<GarbageAnalysisResult> result = douBaoController.analyzeGarbage(param2);
|
||||
if (result.getCode() == 200) {
|
||||
GarbageAnalysisResult data = result.getData();
|
||||
System.out.println("车棚状态：" + (data.getIsShedOpened() ? "已打开" : "未打开"));
|
||||
System.out.println("垃圾类型：" + data.getGarbageType());
|
||||
System.out.println("轻物质占比：" + data.getLightMaterialPercentage() + "%");
|
||||
System.out.println("重物质占比：" + data.getHeavyMaterialPercentage() + "%");
|
||||
}
|
||||
```
|
||||
|
||||
#### 方式二：传统方式
|
||||
|
||||
```java
|
||||
GarbageAnalysisParam param = new GarbageAnalysisParam();
|
||||
param.setImageUrls(Arrays.asList(
|
||||
"https://example.com/garbage-truck-1.jpg",
|
||||
"https://example.com/garbage-truck-2.jpg"
|
||||
));
|
||||
param.setOrderSn("ORD20240101001");
|
||||
param.setRemark("装货地点：南京市建邺区，卸货地点：南京市江宁区");
|
||||
|
||||
R<GarbageAnalysisResult> result = douBaoController.analyzeGarbage(param);
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### JavaScript/Axios 示例
|
||||
|
||||
```javascript
|
||||
// 使用 fetch
|
||||
fetch('http://localhost:8080/douBao/analyzeGarbage', {
|
||||
method: 'POST',
|
||||
headers: {
|
||||
'Content-Type': 'application/json',
|
||||
},
|
||||
body: JSON.stringify({
|
||||
imageUrls: [
|
||||
'https://example.com/garbage-truck-1.jpg',
|
||||
'https://example.com/garbage-truck-2.jpg'
|
||||
],
|
||||
orderSn: 'ORD20240101001',
|
||||
remark: '装货地点：南京市建邺区，卸货地点：南京市江宁区'
|
||||
})
|
||||
})
|
||||
.then(response => response.json())
|
||||
.then(data => {
|
||||
if (data.code === 200) {
|
||||
console.log('分析成功：', data.data);
|
||||
console.log('车棚状态：', data.data.isShedOpened ? '已打开' : '未打开');
|
||||
console.log('垃圾类型：', data.data.garbageType);
|
||||
} else {
|
||||
console.error('分析失败：', data.msg);
|
||||
}
|
||||
})
|
||||
.catch(error => console.error('请求错误：', error));
|
||||
|
||||
// 使用 axios
|
||||
import axios from 'axios';
|
||||
|
||||
const analyzeGarbage = async () => {
|
||||
try {
|
||||
const response = await axios.post('http://localhost:8080/douBao/analyzeGarbage', {
|
||||
imageUrls: [
|
||||
'https://example.com/garbage-truck-1.jpg',
|
||||
'https://example.com/garbage-truck-2.jpg'
|
||||
],
|
||||
orderSn: 'ORD20240101001',
|
||||
remark: '装货地点：南京市建邺区，卸货地点：南京市江宁区'
|
||||
});
|
||||
|
||||
if (response.data.code === 200) {
|
||||
console.log('分析结果：', response.data.data);
|
||||
}
|
||||
} catch (error) {
|
||||
console.error('请求失败：', error);
|
||||
}
|
||||
};
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### Postman/Apifox 测试配置
|
||||
|
||||
#### 请求配置
|
||||
- **请求方式：** `POST`
|
||||
- **请求URL：** `http://localhost:8080/douBao/analyzeGarbage`
|
||||
- **Headers：**
|
||||
- `Content-Type: application/json`
|
||||
|
||||
#### Body（raw JSON）示例
|
||||
|
||||
**示例1：最小参数**
|
||||
```json
|
||||
{
|
||||
"imageUrls": [
|
||||
"https://example.com/garbage-truck-1.jpg"
|
||||
]
|
||||
}
|
||||
```
|
||||
|
||||
**示例2：完整参数**
|
||||
```json
|
||||
{
|
||||
"imageUrls": [
|
||||
"https://example.com/garbage-truck-front.jpg",
|
||||
"https://example.com/garbage-truck-side.jpg",
|
||||
"https://example.com/garbage-truck-back.jpg"
|
||||
],
|
||||
"orderSn": "ORD20240101001",
|
||||
"remark": "装货地点：南京市建邺区奥体中心；卸货地点：南京市江宁区垃圾处理厂"
|
||||
}
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### Python 示例
|
||||
|
||||
```python
|
||||
import requests
|
||||
import json
|
||||
|
||||
url = "http://localhost:8080/douBao/analyzeGarbage"
|
||||
|
||||
# 最小参数
|
||||
payload1 = {
|
||||
"imageUrls": [
|
||||
"https://example.com/garbage-truck-1.jpg"
|
||||
]
|
||||
}
|
||||
|
||||
# 完整参数
|
||||
payload2 = {
|
||||
"imageUrls": [
|
||||
"https://example.com/garbage-truck-front.jpg",
|
||||
"https://example.com/garbage-truck-side.jpg"
|
||||
],
|
||||
"orderSn": "ORD20240101001",
|
||||
"remark": "装货地点：南京市建邺区，卸货地点：南京市江宁区"
|
||||
}
|
||||
|
||||
headers = {
|
||||
"Content-Type": "application/json"
|
||||
}
|
||||
|
||||
response = requests.post(url, headers=headers, data=json.dumps(payload2))
|
||||
result = response.json()
|
||||
|
||||
if result.get("code") == 200:
|
||||
data = result.get("data")
|
||||
print(f"车棚状态：{'已打开' if data.get('isShedOpened') else '未打开'}")
|
||||
print(f"垃圾类型：{data.get('garbageType')}")
|
||||
print(f"轻物质占比：{data.get('lightMaterialPercentage')}%")
|
||||
print(f"重物质占比：{data.get('heavyMaterialPercentage')}%")
|
||||
else:
|
||||
print(f"分析失败：{result.get('msg')}")
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### 测试图片URL说明
|
||||
|
||||
**注意：** 实际测试时，请替换为真实可访问的图片URL。图片URL需要满足以下条件：
|
||||
|
||||
1. **可公开访问**：图片URL必须可以通过HTTP/HTTPS协议直接访问
|
||||
2. **支持格式**：建议使用 JPG、PNG 等常见图片格式
|
||||
3. **图片质量**：建议图片清晰，能够清楚看到车辆和垃圾内容
|
||||
4. **推荐来源**：
|
||||
- OSS对象存储（如阿里云OSS、腾讯云COS等）
|
||||
- CDN加速域名
|
||||
- 公网可访问的图片服务器
|
||||
|
||||
**测试用占位图片（仅供参考，实际使用时请替换）：**
|
||||
```
|
||||
https://via.placeholder.com/800x600.jpg?text=Garbage+Truck+Front
|
||||
https://via.placeholder.com/800x600.jpg?text=Garbage+Truck+Side
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### 本地开发环境测试
|
||||
|
||||
如果是在本地开发环境测试，请确保：
|
||||
|
||||
1. **服务已启动**：确保Spring Boot应用已启动
|
||||
2. **端口正确**：默认端口通常是 `8080`，请根据实际配置调整
|
||||
3. **跨域配置**：如果前端调用，可能需要配置CORS
|
||||
4. **完整URL示例**：
|
||||
- 本地：`http://localhost:8080/douBao/analyzeGarbage`
|
||||
- 开发环境：`http://dev.example.com:8080/douBao/analyzeGarbage`
|
||||
- 生产环境：`https://api.example.com/douBao/analyzeGarbage`
|
||||
|
||||
---
|
||||
|
||||
## 响应示例
|
||||
|
||||
### 成功响应（车棚已打开）
|
||||
|
||||
```json
|
||||
{
|
||||
"code": 200,
|
||||
"msg": "操作成功",
|
||||
"data": {
|
||||
"isShedOpened": true,
|
||||
"details": "石块(多),木材(少)",
|
||||
"lightMaterialPercentage": 10.00,
|
||||
"heavyMaterialPercentage": 90.00,
|
||||
"garbageType": "拆除垃圾",
|
||||
"judgmentBasis": "车厢内主要为石块,这类垃圾多来自建筑物拆除等工程",
|
||||
"analysisTime": 1704067200000,
|
||||
"rawResponse": "明细:石块(多),木材(少);轻物质:10%;重物质:90%;垃圾类型:拆除垃圾;判断依据:车厢内主要为石块,这类垃圾多来自建筑物拆除等工程",
|
||||
"garbageTypes": null,
|
||||
"totalWeight": null
|
||||
}
|
||||
}
|
||||
```
|
||||
|
||||
### 成功响应（车棚未打开）
|
||||
|
||||
```json
|
||||
{
|
||||
"code": 200,
|
||||
"msg": "操作成功",
|
||||
"data": {
|
||||
"isShedOpened": false,
|
||||
"details": "未打开车棚",
|
||||
"lightMaterialPercentage": null,
|
||||
"heavyMaterialPercentage": null,
|
||||
"garbageType": null,
|
||||
"judgmentBasis": null,
|
||||
"analysisTime": 1704067200000,
|
||||
"rawResponse": "未打开车棚",
|
||||
"garbageTypes": null,
|
||||
"totalWeight": null
|
||||
}
|
||||
}
|
||||
```
|
||||
|
||||
### 错误响应
|
||||
|
||||
```json
|
||||
{
|
||||
"code": 500,
|
||||
"msg": "垃圾分析失败：垃圾分析参数不能为空，且必须包含至少一张图片URL",
|
||||
"data": null
|
||||
}
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 错误码说明
|
||||
|
||||
| 错误码 | 说明 |
|
||||
|--------|------|
|
||||
| 200 | 请求成功 |
|
||||
| 400 | 请求参数错误（如：图片URL列表为空） |
|
||||
| 500 | 服务器内部错误（如：AI服务调用失败、解析失败等） |
|
||||
|
||||
---
|
||||
|
||||
## 业务逻辑说明
|
||||
|
||||
### 分析流程
|
||||
|
||||
1. **车棚状态检测**：AI首先判断车棚是否完全打开
|
||||
- 如果车棚未完全打开，直接返回"未打开车棚"，不进行后续分析
|
||||
- 如果车棚已打开，继续执行垃圾分析
|
||||
|
||||
2. **垃圾识别**（仅当车棚已打开时）：
|
||||
- 识别车厢内包含的物品类型和数量
|
||||
- 如果有多辆车，只分析画面中间的车辆
|
||||
- 分析轻物质和重物质的占比（百分比，0-100，两者之和应为100%）
|
||||
- 判断垃圾类型：拆除垃圾、装修垃圾或抛货
|
||||
- 提供判断依据说明
|
||||
|
||||
### 垃圾类型说明
|
||||
|
||||
- **拆除垃圾**：主要来自建筑物拆除等工程，通常包含大量石块、混凝土块等重物质
|
||||
- **装修垃圾**：主要来自室内装修，通常包含白色板材、泡沫、木棍等轻物质
|
||||
- **抛货**：其他类型的垃圾
|
||||
|
||||
---
|
||||
|
||||
## 注意事项
|
||||
|
||||
1. **图片要求**：
|
||||
- 图片URL必须可公开访问
|
||||
- 支持HTTP/HTTPS协议
|
||||
- 建议图片清晰，能够清楚看到车辆和垃圾内容
|
||||
- 支持多张图片，建议从不同角度拍摄
|
||||
|
||||
2. **性能考虑**：
|
||||
- AI分析需要一定时间，建议设置合理的超时时间
|
||||
- 图片数量越多，分析时间可能越长
|
||||
|
||||
3. **数据准确性**：
|
||||
- AI分析结果仅供参考，建议结合人工审核
|
||||
- 如果分析结果不准确，可以通过 `rawResponse` 字段查看AI原始响应进行排查
|
||||
|
||||
4. **字段说明**：
|
||||
- `garbageTypes` 和 `totalWeight` 为保留字段，当前版本可能为null
|
||||
- `isShedOpened` 为false时，其他分析字段（如 `garbageType`、`lightMaterialPercentage` 等）可能为null
|
||||
|
||||
5. **格式要求**：
|
||||
- AI返回的格式为固定格式，解析失败时会记录警告日志
|
||||
- 如果解析结果不完整，建议查看 `rawResponse` 字段获取原始响应
|
||||
|
||||
---
|
||||
|
||||
## 更新日志
|
||||
|
||||
| 版本 | 日期 | 更新内容 |
|
||||
|------|------|----------|
|
||||
| 1.0.0 | 2024-01-01 | 初始版本，支持垃圾车辆图片分析功能 |
|
||||
|
||||
---
|
||||
|
||||
## 联系方式
|
||||
|
||||
如有问题或建议，请联系开发团队。
|
||||
|
||||
|
|
@ -0,0 +1,48 @@
|
|||
package com.njzscloud.supervisory.doubao.config;
|
||||
|
||||
import com.volcengine.ark.runtime.service.ArkService;
|
||||
import lombok.RequiredArgsConstructor;
|
||||
import okhttp3.ConnectionPool;
|
||||
import okhttp3.Dispatcher;
|
||||
import org.springframework.boot.autoconfigure.condition.ConditionalOnMissingBean;
|
||||
import org.springframework.boot.context.properties.EnableConfigurationProperties;
|
||||
import org.springframework.context.annotation.Bean;
|
||||
import org.springframework.context.annotation.Configuration;
|
||||
import org.springframework.util.StringUtils;
|
||||
|
||||
import java.util.concurrent.TimeUnit;
|
||||
|
||||
/**
|
||||
* 火山引擎豆包配置类
|
||||
*/
|
||||
@Configuration
|
||||
@RequiredArgsConstructor
|
||||
@EnableConfigurationProperties(DouBaoProperties.class)
|
||||
public class DouBaoConfiguration {
|
||||
|
||||
private final DouBaoProperties properties;
|
||||
|
||||
@Bean(destroyMethod = "shutdownExecutor")
|
||||
@ConditionalOnMissingBean
|
||||
public ArkService arkService() {
|
||||
String apiKey = properties.getApiKey();
|
||||
if (!StringUtils.hasText(apiKey)) {
|
||||
throw new IllegalStateException("豆包API Key未配置，请设置环境变量 ARK_API_KEY 或在配置文件中设置 doubao.api-key");
|
||||
}
|
||||
|
||||
ConnectionPool connectionPool = new ConnectionPool(
|
||||
properties.getConnectionPoolSize(),
|
||||
properties.getConnectionKeepAliveSeconds(),
|
||||
TimeUnit.SECONDS
|
||||
);
|
||||
Dispatcher dispatcher = new Dispatcher();
|
||||
|
||||
return ArkService.builder()
|
||||
.dispatcher(dispatcher)
|
||||
.connectionPool(connectionPool)
|
||||
.baseUrl(properties.getBaseUrl())
|
||||
.apiKey(apiKey)
|
||||
.build();
|
||||
}
|
||||
}
|
||||
|
||||
|
|
@ -0,0 +1,52 @@
|
|||
package com.njzscloud.supervisory.doubao.config;
|
||||
|
||||
import lombok.Getter;
|
||||
import lombok.Setter;
|
||||
import lombok.ToString;
|
||||
import lombok.experimental.Accessors;
|
||||
import org.springframework.boot.context.properties.ConfigurationProperties;
|
||||
|
||||
/**
|
||||
* 火山引擎豆包配置
|
||||
*/
|
||||
@Getter
|
||||
@Setter
|
||||
@ToString
|
||||
@Accessors(chain = true)
|
||||
@ConfigurationProperties(prefix = "doubao")
|
||||
public class DouBaoProperties {
|
||||
|
||||
/**
|
||||
* API Key，优先从环境变量 ARK_API_KEY 获取，如果环境变量不存在则使用配置值
|
||||
*/
|
||||
private String apiKey;
|
||||
|
||||
/**
|
||||
* 基础URL，默认为 https://ark.cn-beijing.volces.com/api/v3
|
||||
*/
|
||||
private String baseUrl = "https://ark.cn-beijing.volces.com/api/v3";
|
||||
|
||||
/**
|
||||
* 模型ID，默认为 doubao-seed-1-6-vision-250815
|
||||
*/
|
||||
private String model = "doubao-seed-1-6-vision-250815";
|
||||
|
||||
/**
|
||||
* 连接池大小，默认5
|
||||
*/
|
||||
private Integer connectionPoolSize = 5;
|
||||
|
||||
/**
|
||||
* 连接保持时间（秒），默认1
|
||||
*/
|
||||
private Long connectionKeepAliveSeconds = 1L;
|
||||
|
||||
/**
|
||||
* 获取API Key，优先从环境变量获取
|
||||
*/
|
||||
public String getApiKey() {
|
||||
String envApiKey = System.getenv("ARK_API_KEY");
|
||||
return envApiKey != null ? envApiKey : apiKey;
|
||||
}
|
||||
}
|
||||
|
||||
|
|
@ -0,0 +1,31 @@
|
|||
package com.njzscloud.supervisory.doubao.controller;
|
||||
|
||||
import com.njzscloud.common.core.utils.R;
|
||||
import com.njzscloud.supervisory.doubao.param.GarbageAnalysisParam;
|
||||
import com.njzscloud.supervisory.doubao.result.GarbageAnalysisResult;
|
||||
import com.njzscloud.supervisory.doubao.sevice.DouBaoService;
|
||||
import lombok.RequiredArgsConstructor;
|
||||
import lombok.extern.slf4j.Slf4j;
|
||||
import org.springframework.web.bind.annotation.PostMapping;
|
||||
import org.springframework.web.bind.annotation.RequestBody;
|
||||
import org.springframework.web.bind.annotation.RequestMapping;
|
||||
import org.springframework.web.bind.annotation.RestController;
|
||||
|
||||
/**
|
||||
* 豆包
|
||||
* 使用豆包官方SDK的生产级实现
|
||||
*/
|
||||
@Slf4j
|
||||
@RestController
|
||||
@RequestMapping("/douBao")
|
||||
@RequiredArgsConstructor
|
||||
public class DouBaoController {
|
||||
|
||||
private final DouBaoService douBaoService;
|
||||
|
||||
@PostMapping("/analyzeGarbage")
|
||||
public R<GarbageAnalysisResult> analyzeGarbage(@RequestBody GarbageAnalysisParam param) {
|
||||
return R.success(douBaoService.analyzeGarbage(param));
|
||||
}
|
||||
|
||||
}
|
||||
|
|
@ -0,0 +1,34 @@
|
|||
package com.njzscloud.supervisory.doubao.param;
|
||||
|
||||
import lombok.Getter;
|
||||
import lombok.Setter;
|
||||
import lombok.ToString;
|
||||
import lombok.experimental.Accessors;
|
||||
|
||||
import java.util.List;
|
||||
|
||||
/**
|
||||
* 垃圾分析请求参数
|
||||
*/
|
||||
@Getter
|
||||
@Setter
|
||||
@ToString
|
||||
@Accessors(chain = true)
|
||||
public class GarbageAnalysisParam {
|
||||
|
||||
/**
|
||||
* 图片URL列表（支持多张图片）
|
||||
*/
|
||||
private List<String> imageUrls;
|
||||
|
||||
/**
|
||||
* 订单号（可选）
|
||||
*/
|
||||
private String orderSn;
|
||||
|
||||
/**
|
||||
* 备注信息（可选）
|
||||
*/
|
||||
private String remark;
|
||||
}
|
||||
|
||||
|
|
@ -0,0 +1,105 @@
|
|||
package com.njzscloud.supervisory.doubao.result;
|
||||
|
||||
import lombok.Getter;
|
||||
import lombok.Setter;
|
||||
import lombok.ToString;
|
||||
import lombok.experimental.Accessors;
|
||||
|
||||
import java.math.BigDecimal;
|
||||
import java.util.List;
|
||||
|
||||
/**
|
||||
* 垃圾分析结果
|
||||
*/
|
||||
@Getter
|
||||
@Setter
|
||||
@ToString
|
||||
@Accessors(chain = true)
|
||||
public class GarbageAnalysisResult {
|
||||
|
||||
/**
|
||||
* 车棚是否完全打开
|
||||
*/
|
||||
private Boolean isShedOpened;
|
||||
|
||||
/**
|
||||
* 明细：车厢内包含的物品列表，格式：物品1(数量描述),物品2(数量描述)
|
||||
* 例如：石块(多),木材(少)
|
||||
*/
|
||||
private String details;
|
||||
|
||||
/**
|
||||
* 轻物质占比（百分比，0-100）
|
||||
*/
|
||||
private BigDecimal lightMaterialPercentage;
|
||||
|
||||
/**
|
||||
* 重物质占比（百分比，0-100）
|
||||
*/
|
||||
private BigDecimal heavyMaterialPercentage;
|
||||
|
||||
/**
|
||||
* 垃圾类型：拆除垃圾、装修垃圾、抛货
|
||||
*/
|
||||
private String garbageType;
|
||||
|
||||
/**
|
||||
* 判断依据：说明判断垃圾类型的原因
|
||||
*/
|
||||
private String judgmentBasis;
|
||||
|
||||
/**
|
||||
* 分析时间戳
|
||||
*/
|
||||
private Long analysisTime;
|
||||
|
||||
/**
|
||||
* 原始响应内容
|
||||
*/
|
||||
private String rawResponse;
|
||||
|
||||
/**
|
||||
* 垃圾类型分析列表（保留用于兼容）
|
||||
*/
|
||||
private List<GarbageTypeInfo> garbageTypes;
|
||||
|
||||
/**
|
||||
* 总重量（单位：千克）（保留用于兼容）
|
||||
*/
|
||||
private BigDecimal totalWeight;
|
||||
|
||||
/**
|
||||
* 垃圾类型信息（保留用于兼容）
|
||||
*/
|
||||
@Getter
|
||||
@Setter
|
||||
@ToString
|
||||
@Accessors(chain = true)
|
||||
public static class GarbageTypeInfo {
|
||||
/**
|
||||
* 垃圾类型名称（如：可回收垃圾、有害垃圾、厨余垃圾、其他垃圾等）
|
||||
*/
|
||||
private String typeName;
|
||||
|
||||
/**
|
||||
* 具体垃圾种类（如：塑料瓶、废纸、电池等）
|
||||
*/
|
||||
private String category;
|
||||
|
||||
/**
|
||||
* 判断依据（AI识别该垃圾类型的理由）
|
||||
*/
|
||||
private String judgmentBasis;
|
||||
|
||||
/**
|
||||
* 占比（百分比，0-100）
|
||||
*/
|
||||
private BigDecimal percentage;
|
||||
|
||||
/**
|
||||
* 估算重量（单位：千克）
|
||||
*/
|
||||
private BigDecimal weight;
|
||||
}
|
||||
}
|
||||
|
||||
|
|
@ -0,0 +1,18 @@
|
|||
package com.njzscloud.supervisory.doubao.sevice;
|
||||
|
||||
import com.njzscloud.supervisory.doubao.param.GarbageAnalysisParam;
|
||||
import com.njzscloud.supervisory.doubao.result.GarbageAnalysisResult;
|
||||
|
||||
/**
|
||||
* 火山引擎豆包服务接口
|
||||
*/
|
||||
public interface DouBaoService {
|
||||
|
||||
/**
|
||||
* 垃圾分析 - 分析多张图片中的垃圾类型、占比和重量
|
||||
*
|
||||
* @param param 垃圾分析请求参数，包含多张图片URL
|
||||
* @return 垃圾分析结果，包含垃圾类型、判断依据、占比、重量等信息
|
||||
*/
|
||||
GarbageAnalysisResult analyzeGarbage(GarbageAnalysisParam param);
|
||||
}
|
||||
|
|
@ -0,0 +1,242 @@
|
|||
package com.njzscloud.supervisory.doubao.sevice.impl;
|
||||
|
||||
import com.njzscloud.supervisory.doubao.config.DouBaoProperties;
|
||||
import com.njzscloud.supervisory.doubao.param.GarbageAnalysisParam;
|
||||
import com.njzscloud.supervisory.doubao.result.GarbageAnalysisResult;
|
||||
import com.njzscloud.supervisory.doubao.sevice.DouBaoService;
|
||||
import com.volcengine.ark.runtime.model.completion.chat.ChatCompletionContentPart;
|
||||
import com.volcengine.ark.runtime.model.completion.chat.ChatCompletionRequest;
|
||||
import com.volcengine.ark.runtime.model.completion.chat.ChatMessage;
|
||||
import com.volcengine.ark.runtime.model.completion.chat.ChatMessageRole;
|
||||
import com.volcengine.ark.runtime.service.ArkService;
|
||||
import lombok.RequiredArgsConstructor;
|
||||
import lombok.extern.slf4j.Slf4j;
|
||||
import org.springframework.stereotype.Service;
|
||||
import org.springframework.util.CollectionUtils;
|
||||
import org.springframework.util.StringUtils;
|
||||
|
||||
import java.math.BigDecimal;
|
||||
import java.util.ArrayList;
|
||||
import java.util.List;
|
||||
import java.util.regex.Matcher;
|
||||
import java.util.regex.Pattern;
|
||||
import java.util.stream.Collectors;
|
||||
|
||||
/**
|
||||
* 火山引擎豆包服务实现类
|
||||
*/
|
||||
@Slf4j
|
||||
@Service
|
||||
@RequiredArgsConstructor
|
||||
public class DouBaoServiceImpl implements DouBaoService {
|
||||
|
||||
private final ArkService arkService;
|
||||
private final DouBaoProperties douBaoProperties;
|
||||
|
||||
/**
|
||||
* 垃圾分析 - 分析多张图片中的垃圾类型、占比和重量
|
||||
*
|
||||
* @param param 垃圾分析请求参数，包含多张图片URL
|
||||
* @return 垃圾分析结果，包含垃圾类型、判断依据、占比、重量等信息
|
||||
*/
|
||||
@Override
|
||||
public GarbageAnalysisResult analyzeGarbage(GarbageAnalysisParam param) {
|
||||
try {
|
||||
log.info("开始调用火山引擎豆包API进行垃圾分析，订单号：{}，图片数量：{}",
|
||||
param != null ? param.getOrderSn() : "未知",
|
||||
param != null && param.getImageUrls() != null ? param.getImageUrls().size() : 0);
|
||||
|
||||
// 参数校验
|
||||
if (param == null || CollectionUtils.isEmpty(param.getImageUrls())) {
|
||||
throw new IllegalArgumentException("垃圾分析参数不能为空，且必须包含至少一张图片URL");
|
||||
}
|
||||
|
||||
// 构建消息列表
|
||||
final List<ChatMessage> messages = new ArrayList<>();
|
||||
final List<ChatCompletionContentPart> multiParts = new ArrayList<>();
|
||||
|
||||
// 添加所有图片URL
|
||||
for (String imageUrl : param.getImageUrls()) {
|
||||
if (StringUtils.hasText(imageUrl)) {
|
||||
multiParts.add(ChatCompletionContentPart.builder()
|
||||
.type("image_url")
|
||||
.imageUrl(new ChatCompletionContentPart.ChatCompletionContentPartImageURL(imageUrl))
|
||||
.build());
|
||||
}
|
||||
}
|
||||
|
||||
// 添加文本提示词，要求返回JSON格式
|
||||
String textPrompt = buildGarbageAnalysisPrompt(param);
|
||||
multiParts.add(ChatCompletionContentPart.builder()
|
||||
.type("text")
|
||||
.text(textPrompt)
|
||||
.build());
|
||||
|
||||
// 构建用户消息
|
||||
final ChatMessage userMessage = ChatMessage.builder()
|
||||
.role(ChatMessageRole.USER)
|
||||
.multiContent(multiParts)
|
||||
.build();
|
||||
messages.add(userMessage);
|
||||
|
||||
// 构建请求
|
||||
ChatCompletionRequest chatCompletionRequest = ChatCompletionRequest.builder()
|
||||
.model(douBaoProperties.getModel())
|
||||
.messages(messages)
|
||||
.build();
|
||||
|
||||
// 调用API并获取响应
|
||||
String response = arkService.createChatCompletion(chatCompletionRequest)
|
||||
.getChoices()
|
||||
.stream()
|
||||
.map(choice -> String.valueOf(choice.getMessage().getContent()))
|
||||
.collect(Collectors.joining("\n"));
|
||||
|
||||
log.info("火山引擎豆包API调用成功，原始响应内容：{}", response);
|
||||
|
||||
// 解析响应并转换为结果对象
|
||||
GarbageAnalysisResult result = parseGarbageAnalysisResponse(response);
|
||||
result.setRawResponse(response);
|
||||
result.setAnalysisTime(System.currentTimeMillis());
|
||||
|
||||
log.info("垃圾分析完成，车棚状态：{}，垃圾类型：{}，轻物质占比：{}%，重物质占比：{}%",
|
||||
result.getIsShedOpened() != null && result.getIsShedOpened() ? "已打开" : "未打开",
|
||||
result.getGarbageType(),
|
||||
result.getLightMaterialPercentage(),
|
||||
result.getHeavyMaterialPercentage());
|
||||
|
||||
return result;
|
||||
|
||||
} catch (Exception e) {
|
||||
log.error("调用火山引擎豆包API进行垃圾分析失败，订单号：{}",
|
||||
param != null ? param.getOrderSn() : "未知", e);
|
||||
throw new RuntimeException("垃圾分析失败：" + e.getMessage(), e);
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* 构建垃圾分析提示词
|
||||
* 要求AI返回指定格式的文本
|
||||
*/
|
||||
private String buildGarbageAnalysisPrompt(GarbageAnalysisParam param) {
|
||||
StringBuilder prompt = new StringBuilder();
|
||||
prompt.append("请仔细分析这些图片中的垃圾内容。\n\n");
|
||||
|
||||
// 添加订单信息（如果有）
|
||||
if (StringUtils.hasText(param.getOrderSn())) {
|
||||
prompt.append("订单号：").append(param.getOrderSn()).append("\n");
|
||||
}
|
||||
if (StringUtils.hasText(param.getRemark())) {
|
||||
prompt.append("备注：").append(param.getRemark()).append("\n");
|
||||
}
|
||||
|
||||
prompt.append("\n请按照以下要求进行分析：\n");
|
||||
prompt.append("1. 首先判断图片中是否有垃圾运输车辆：\n");
|
||||
prompt.append(" - 如果图片中没有车辆（直接是垃圾堆、垃圾场等场景），请直接跳到第3步分析垃圾内容\n");
|
||||
prompt.append(" - 如果图片中有车辆，请继续第2步判断车棚状态\n");
|
||||
prompt.append("2. 如果图片中有车辆，判断车棚是否完全打开：\n");
|
||||
prompt.append(" - 如果车棚没有完全打开，无法看到车厢内的垃圾内容，直接回答：未打开车棚\n");
|
||||
prompt.append(" - 如果车棚完全打开，可以看到车厢内的垃圾，请继续第3步分析\n");
|
||||
prompt.append("3. 分析垃圾内容（适用于：无车辆的场景，或有车辆且车棚已打开的场景）：\n");
|
||||
prompt.append(" - 识别垃圾中包含哪几种物品（如果有多辆车，只分析画面中间的车辆或主要垃圾堆）\n");
|
||||
prompt.append(" - 说明哪种物品占比最多\n");
|
||||
prompt.append(" - 给出轻物质和重物质的大概占比（使用百分比数值，0-100）\n");
|
||||
prompt.append(" - 判断垃圾类型：拆除垃圾、装修垃圾或抛货\n");
|
||||
prompt.append(" - 说明判断垃圾类型的原因\n\n");
|
||||
|
||||
prompt.append("请严格按照以下格式返回结果，不要添加任何其他文字说明：\n");
|
||||
prompt.append("如果图片中有车辆且车棚未打开，直接回答：未打开车棚\n");
|
||||
prompt.append("如果图片中没有车辆，或者有车辆且车棚已打开，按照以下格式回答：\n");
|
||||
prompt.append("明细:物品1(数量描述),物品2(数量描述);轻物质:x%;重物质:x%;垃圾类型:x;判断依据:x\n\n");
|
||||
prompt.append("格式说明：\n");
|
||||
prompt.append("- 明细：列出垃圾中包含的所有物品，用逗号分隔，每个物品后标注数量（多、少、较多、较少等）\n");
|
||||
prompt.append("- 轻物质：轻物质的占比百分比（0-100）\n");
|
||||
prompt.append("- 重物质：重物质的占比百分比（0-100），轻物质和重物质占比之和应为100%\n");
|
||||
prompt.append("- 垃圾类型：拆除垃圾、装修垃圾或抛货\n");
|
||||
prompt.append("- 判断依据：详细说明为什么判断为该垃圾类型\n\n");
|
||||
prompt.append("示例：\n");
|
||||
prompt.append("明细:石块(多);轻物质:0%;重物质:100%;垃圾类型:拆除垃圾;判断依据:主要为石块,这类垃圾多来自建筑物拆除等工程\n");
|
||||
prompt.append("明细:白色板材(多),白色泡沫(少),木棍(少);轻物质:90%;重物质:10%;垃圾类型:装修垃圾;判断依据:主要是白色板材、泡沫等装修常见废弃物\n");
|
||||
prompt.append("明细:纸箱(多),塑料瓶(少),泡沫箱(少);轻物质:95%;重物质:5%;垃圾类型:抛货;判断依据:主要是轻质包装材料,属于抛货类垃圾\n\n");
|
||||
prompt.append("注意：必须严格按照上述格式返回，不要添加任何其他文字说明。");
|
||||
|
||||
return prompt.toString();
|
||||
}
|
||||
|
||||
/**
|
||||
* 解析垃圾分析响应
|
||||
* 从AI响应中提取信息并转换为结果对象
|
||||
* 格式：明细:xx(x),xx(x);轻物质:x%;重物质:x%;垃圾类型:x;判断依据:x
|
||||
* 或者：未打开车棚
|
||||
*/
|
||||
private GarbageAnalysisResult parseGarbageAnalysisResponse(String response) {
|
||||
if (!StringUtils.hasText(response)) {
|
||||
throw new RuntimeException("AI响应为空，无法解析");
|
||||
}
|
||||
|
||||
try {
|
||||
GarbageAnalysisResult result = new GarbageAnalysisResult();
|
||||
String trimmed = response.trim();
|
||||
|
||||
// 检查是否是"未打开车棚"的情况
|
||||
if (trimmed.contains("未打开车棚")) {
|
||||
result.setIsShedOpened(false);
|
||||
result.setDetails("未打开车棚");
|
||||
result.setLightMaterialPercentage(BigDecimal.ZERO);
|
||||
result.setHeavyMaterialPercentage(BigDecimal.ZERO);
|
||||
result.setGarbageType("无");
|
||||
result.setJudgmentBasis("车棚未完全打开，无法进行分析");
|
||||
return result;
|
||||
}
|
||||
|
||||
// 车棚已打开，解析详细格式
|
||||
result.setIsShedOpened(true);
|
||||
|
||||
// 提取明细
|
||||
Pattern detailsPattern = Pattern.compile("明细[:：]([^;]+)");
|
||||
Matcher detailsMatcher = detailsPattern.matcher(trimmed);
|
||||
if (detailsMatcher.find()) {
|
||||
result.setDetails(detailsMatcher.group(1).trim());
|
||||
}
|
||||
|
||||
// 提取轻物质占比
|
||||
Pattern lightPattern = Pattern.compile("轻物质[:：](\\d+(?:\\.\\d+)?)%");
|
||||
Matcher lightMatcher = lightPattern.matcher(trimmed);
|
||||
if (lightMatcher.find()) {
|
||||
result.setLightMaterialPercentage(new BigDecimal(lightMatcher.group(1)));
|
||||
}
|
||||
|
||||
// 提取重物质占比
|
||||
Pattern heavyPattern = Pattern.compile("重物质[:：](\\d+(?:\\.\\d+)?)%");
|
||||
Matcher heavyMatcher = heavyPattern.matcher(trimmed);
|
||||
if (heavyMatcher.find()) {
|
||||
result.setHeavyMaterialPercentage(new BigDecimal(heavyMatcher.group(1)));
|
||||
}
|
||||
|
||||
// 提取垃圾类型
|
||||
Pattern typePattern = Pattern.compile("垃圾类型[:：]([^;]+)");
|
||||
Matcher typeMatcher = typePattern.matcher(trimmed);
|
||||
if (typeMatcher.find()) {
|
||||
result.setGarbageType(typeMatcher.group(1).trim());
|
||||
}
|
||||
|
||||
// 提取判断依据
|
||||
Pattern basisPattern = Pattern.compile("判断依据[:：](.+)");
|
||||
Matcher basisMatcher = basisPattern.matcher(trimmed);
|
||||
if (basisMatcher.find()) {
|
||||
result.setJudgmentBasis(basisMatcher.group(1).trim());
|
||||
}
|
||||
|
||||
// 验证必要字段
|
||||
if (result.getDetails() == null || result.getGarbageType() == null) {
|
||||
log.warn("解析结果不完整，响应内容：{}", response);
|
||||
}
|
||||
|
||||
return result;
|
||||
|
||||
} catch (Exception e) {
|
||||
log.error("解析垃圾分析响应失败，响应内容：{}", response, e);
|
||||
throw new RuntimeException("解析AI响应失败：" + e.getMessage(), e);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
|
@ -80,6 +80,9 @@ wechat:
|
|||
notify-url: http://115.29.236.92:8082/payment/wechat/notify
|
||||
# 退款回调地址
|
||||
refund-notify-url: http://115.29.236.92:8082/payment/wechat/refundNotify
|
||||
doubao:
|
||||
api-name: api-key-20250411103646
|
||||
api-key: e0fc97bb-2bb6-4525-a79b-1e3555026821
|
||||
mqtt:
|
||||
enabled: true
|
||||
broker: tcp://139.224.54.144:1883
|
||||
|
|
|
|||
|
|
@ -70,7 +70,9 @@ wechat:
|
|||
notify-url: http://139.224.32.69:80/api/payment/wechat/notify
|
||||
# 退款回调地址
|
||||
refund-notify-url: http://139.224.32.69:80/api/payment/wechat/refundNotify
|
||||
|
||||
doubao:
|
||||
api-name: api-key-20250411103646
|
||||
api-key: e0fc97bb-2bb6-4525-a79b-1e3555026821
|
||||
mqtt:
|
||||
enabled: true
|
||||
broker: tcp://127.0.0.1:1883
|
||||
|
|
|
|||
Loading…
Reference in New Issue