wadhwani-ai/pest-management-opendata

---
license: apache-2.0
---

# Wadhwani AI Pest Management Open Data

This dataset is a Hugging Face adaptor to the official dataset [hosted
on
Github](https://github.com/wadhwani-ai/pest-management-opendata). Please
refer to that repository for detailed and up-to-date documentation.

## Usage

This dataset is large. It is strongly recommended users access it as a
stream:

```python
from datasets import load_dataset
dataset = load_dataset('wadhwani-ai/pest-management-opendata', streaming=True)
```

Bounding boxes are stored as geospatial types. Once loaded, they can be
read as follows:

```python
from shapely.wkb import loads

for (s, data) in dataset.items():
for d in data:
pests = d['pests']
iterable = map(pests.get, ('label', 'geometry'))
for (i, j) in zip(*iterable):
geom = loads(j)
print(i, geom.bounds)
```

The bounds of a geometry are what most object detection systems
require. See the [Shapely
documentation](https://shapely.readthedocs.io/en/stable/manual.html#object.bounds)
for more.

原始信息汇总

Wadhwani AI Pest Management Open Data

使用说明

该数据集较大，强烈建议用户以流式方式访问：

python from datasets import load_dataset dataset = load_dataset(wadhwani-ai/pest-management-opendata, streaming=True)

边界框以地理空间类型存储。加载后，可以按以下方式读取：

python from shapely.wkb import loads

for (s, data) in dataset.items(): for d in data: pests = d[pests] iterable = map(pests.get, (label, geometry)) for (i, j) in zip(*iterable): geom = loads(j) print(i, geom.bounds)

几何边界是大多数目标检测系统所需的内容。更多信息请参阅 Shapely 文档。