ยูโรแซท

  • คำอธิบาย :

ชุดข้อมูล EuroSAT อ้างอิงจากภาพถ่ายดาวเทียม Sentinel-2 ซึ่งครอบคลุม 13 แถบสเปกตรัมและประกอบด้วย 10 คลาสพร้อมตัวอย่างที่มีป้ายกำกับและอ้างอิงทางภูมิศาสตร์ 27,000 รายการ

มีชุดข้อมูลสองชุด: - rgb: มีเฉพาะแถบความถี่ออปติคอล R, G, B ที่เข้ารหัสเป็นภาพ JPEG - ทั้งหมด: มีทั้งหมด 13 แบนด์ในช่วงค่าดั้งเดิม (float32)

URL: https://github.com/phelber/eurosat

แยก ตัวอย่าง
'train' 27,000
  • การอ้างอิง :
@misc{helber2017eurosat,
    title={EuroSAT: A Novel Dataset and Deep Learning Benchmark for Land Use and Land Cover Classification},
    author={Patrick Helber and Benjamin Bischke and Andreas Dengel and Damian Borth},
    year={2017},
    eprint={1709.00029},
    archivePrefix={arXiv},
    primaryClass={cs.CV}
}

eurosat/rgb (การกำหนดค่าเริ่มต้น)

  • คำอธิบาย การกำหนดค่า: ช่องสัญญาณ RGB ของ Sentinel-2

  • ขนาดการดาวน์โหลด : 89.91 MiB

  • ขนาดชุดข้อมูล : 89.50 MiB

  • แคชอัตโนมัติ ( เอกสาร ): ใช่

  • โครงสร้างคุณลักษณะ :

FeaturesDict({
    'filename': Text(shape=(), dtype=string),
    'image': Image(shape=(64, 64, 3), dtype=uint8),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=10),
})
  • เอกสารคุณสมบัติ :
ลักษณะเฉพาะ ระดับ รูปร่าง Dประเภท คำอธิบาย
คุณสมบัติDict
ชื่อไฟล์ ข้อความ สตริง
ภาพ ภาพ (64, 64, 3) uint8
ฉลาก ป้ายกำกับคลาส int64

การสร้างภาพ

ยูโรแซท/ทั้งหมด

  • คำอธิบาย การกำหนดค่า : 13 Sentinel-2 ช่องสัญญาณ

  • ขนาดการดาวน์โหลด : 1.93 GiB

  • ขนาดชุดข้อมูล : 5.36 GiB

  • แคชอัตโนมัติ ( เอกสารประกอบ ): ไม่

  • โครงสร้างคุณลักษณะ :

FeaturesDict({
    'filename': Text(shape=(), dtype=string),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=10),
    'sentinel2': Tensor(shape=(64, 64, 13), dtype=float32),
})
  • เอกสารคุณสมบัติ :
ลักษณะเฉพาะ ระดับ รูปร่าง Dประเภท คำอธิบาย
คุณสมบัติDict
ชื่อไฟล์ ข้อความ สตริง
ฉลาก ป้ายกำกับคลาส int64
ยาม2 เทนเซอร์ (64, 64, 13) ลอย32