-
-
-
-
-
-
-
- 
SV248S Dataset:
- Single Objection Tracking Dataset
-Satellite Videos -
-
-
-
-
-
- -
- 该数据集是由吉林一号卫星创建，由分辨率为0.92米的高清全色图构成。 - 数据集使用长光官方开源的6个视频，每个视频选择约40个目标。 - 每个视频有500-753帧，持续20-30.12秒，帧速率为25fps，包括四个国家，五个城市，共248个物体。 - SV中有船舶、车辆、大型车辆和飞机四类对象，主要关注难以跟踪的小目标（车辆、大型车辆，占物体比例的90%）。 - 有关该数据集更详细的数据见“Deep Learning-Based Object Tracking in Satellite Videos: A Comprehensive Survey With a New Dataset”。 - 该数据集由西安电子科技大学IPIU实验室（智能感知与图像理解教育部重点实验室）于2021年7月完成。 -
- -
-
-
-
-
-
- Github
-
-
-
-
-
-
-
-
- Paper
-
-
-
-
-
-
-
-
-
-
-
-
-
- Image
-
-
- 
-
-
-
-
-
-
-
-Tags Instruction Following:
-
-
-
-
-
-| - | Name | -Definition | -Dependence | -AUTO | -Basic Properties | -Center Point(CP) | -It represents the CP of the object bounding box. | -BBox | -Y | - - -
|---|---|---|---|---|
| Movement Velocity(MV) | -It calculates the average velocity considering both the speed value and the direction in adjacent five frames (unit: pps) | -CP | -Y | -|
| Object Size(OS) | -The largest value is selected from the height and width of the object bounding box. | -BBox | -Y | -|
| Frame State Flags | -Invisible(INV) | -The object disappeared without any occluder or is too similar to its surroundings. | -M | -N | -
| Normal Visible(NOR) | -The object is visible and found easily. | -M | -N | -|
| Occlusion(OCC) | -The object is in the shadow of the building or behind something. | -M | -N | -|
| Sequence Attributes | -Background Change(BCH) | -The background of the tracked object has noticeable changes in color or texture. | -M | -N | -
| Background Cluster(BCL) | -There are at least 10 frames that contain in the INV flag. | -INV | -Y | -|
| Continuous Occlusion(CO) | -STO or LTO occur twice or more times in a sequence. | -STO, LTO | -Y | -|
| Dense Similarity(DS) | -One or more similar objects exist around the tracked object in the range of 2.5 times OS. | -OS | -N | -|
| In-plane Rotation(IPR) | -The object has an in-plane rotation at an angle greater than or equal to 30°. | -MV | -Y | -|
| Illumination Variation(IV) | -The object has noticeable changes in brightness or color. | -M | -N | -|
| Short-term Occlusion(STO) | -The sequence exists less than or equal to 50 conseccutive frames with OCC flag. | -OCC | -Y | -|
| Long-term Occlusion(LTO) | -The sequence exists more than 50 consecutive frames with OCC flag. | -OCC | -Y | -|
| Natural Disturbance(ND) | -The object's appearance is influenced by smog or sandy weather or blocked by clouds. | -M | -N | -|
| Slow Motion(SM) | -The moving speed of the tracked object is fewer than 2.2pps. | -MV | -Y | -
-
-
-
-
-
-
-
- Design Principle
- -
- 
-
-
-
-
diff --git a/index.md b/index.md
index 0e1f413..2bef7a7 100644
--- a/index.md
+++ b/index.md
@@ -1,81 +1,14 @@
-
-
- File Definition:
- - - -{
- "source_info":{
- "video_id": "01", # 源视频的名称
- "seq_id": "000002", # 当前目标的名称
- "frame_range": [], # 相对于原始视频使用的帧范围
- "crop_range": [] # 相对于原始视频的裁剪块的范围
- },
- "details":{
- "init_rect": [], # 初始化边界框
- "init_poly": [], # 初始化的紧致多边形
- "length": 325, # 序列的总长度
- "class_name": "ship", # ship, plane, car, car-large (这些名称与论文中不同)
- "level": "simple" # simple, normal, hard
- }
-}SV248S Dataset
+layout: default -人生就是一个不断学习的过程，通过学习来充实自己的头脑，让自己更有智慧的生活下去，去探索那未知领域的神秘，去享受成长带来的快乐和惬意。
+title: My Blog -**Authors:** Yuxuna Li, Licheng Jiao, Zhongjian Huang, Xin Zhang, Ruohan Zhang, Xue Song, Chenxi Tian, Zixiao Zhang, Fang Liu, Shuyuan Yang, Biao Hou, Wenping Ma, Xu Liu, Lingling Li
-School of IPIU Laboratory, Xidian University
-This is a regular paragraph. +--- -| Foo | -