Back to datasets
Dataset assetOpen Source CommunityComputer VisionImage Processing
BSDS500/300, BSD68, Set12
BSDS500/300 is a dataset provided by the Berkeley Vision Lab for image segmentation or contour detection, and is also used for super‑resolution reconstruction. The database contains 200 training images, 200 validation images, and 100 test images, with ground‑truth annotations stored in MAT files. BSD68 is a color dataset for image denoising benchmarks and is part of the Berkeley Segmentation Dataset and Benchmark. Set12 contains 12 images for evaluating image denoising algorithms.
Source
github
Created
May 18, 2019
Updated
Nov 1, 2021
Signals
1,071 views
Availability
Linked source ready
Overview
Dataset description and usage context
Dataset Overview
BSDS500/300
- Source: Provided by the Berkeley Vision Lab.
- Purpose: Primarily for image segmentation and contour detection, also applicable to super‑resolution reconstruction.
- Composition: 200 training images, 200 validation images, 100 test images.
- Annotations: Manually labeled ground truth saved as MAT files, containing annotations from multiple raters.
- File Format: MAT, convenient for direct loading in MATLAB.
Download Links
Loading Scripts
- Contour Visualization: Use
make_gt_bondary_image.mto generate binary contour maps. - Segmentation Visualization: Use
make_gt_seg_image.mto produce segmentation visualizations.
BSD68
- Purpose: Benchmark for image denoising algorithms.
- Download: BSD68 Download
Set12
- Download: Set12 Download
- Features: Contains 12 images.
Need downstream help?
Pair the dataset with AI analysis and content workflows.
Once the source passes your review, move straight into summarization, transformation, report drafting, or presentation generation with the JuheAI toolchain.