Datasets
Standard Dataset
Three Surveillance Test Videos of Hall, PETS, and CW
- Citation Author(s):
- Submitted by:
- Zhen Song
- Last updated:
- Fri, 10/27/2023 - 03:16
- DOI:
- 10.21227/gy2n-rt89
- License:
- Categories:
- Keywords:
Abstract
The dataset consists of six .mat files containing three surveillance video test sequences, Hall_qcif_330 (Hall, 330 frames), PETS2009_S1L1-View_001 (PETS, 100 frames), and Crosswalk (CW, 270 frames), and the corresponding background image for three videos (Only the data of each video's gray channel component). Hall is shot indoors and disturbed by noise, PETS is shot outdoors with less noise, and CW is shot outdoors with heavy noise interference. Hall and PETS are two foreground-sparse videos with small objects. CW is a foreground-dense video with dramatic changes in sparsity. All the video sequences are in a native or trimmed resolution of QCIF.
All the video sequences (only for its gray channel component) are in a native or trimmed resolution of QCIF.
Hall_seq.mat
The test video of Hall_qcif_330, from: https://media.xiph.org/video/derf/
PETS_seq.mat
The test video of PETS2009_S1L1-View_001, provided by J. Ferryman and A. Shahrokni, “Pets2009: Dataset and challenge,” in Proc. 12th Int. Workshop Perform. Eval. Tracking Surveill., Jan. 2009, pp. 1–6.
CW_seq.mat
The test video of Crosswalk, used in J. Wang, W. Wang, and J. Chen, “Adaptive rate block compressive sensing based on statistical characteristics estimation,” IEEE Trans. Image Process., vol. 31, pp. 734–747, 2022.
Hall/PETS/CW_Background.mat
The background images of the corresponding test videos
Dataset Files
- CW_Seq.mat (5.04 MB)
- CW_Background.mat (16.04 kB)
- Hall_Seq.mat (6.43 MB)
- Hall_Background.mat (18.64 kB)
- PETS_Seq.mat (2.07 MB)
- PETS_Background.mat (21.12 kB)
- readme.txt (751 bytes)
Comments
Test Videos of Hall, PETS, and CW
Test Videos of Hall, PETS, and CW
Uploading the test videos of Hall, PETS, and CW, and their backgrounds.
Uploading the test videos of Hall, PETS, and CW, and their backgrounds.