norbertm commited on
Commit
03c6d59
·
verified ·
1 Parent(s): 781466e

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +89 -83
README.md CHANGED
@@ -1,83 +1,89 @@
1
- ---
2
- dataset_info:
3
- features:
4
- - name: id
5
- dtype: string
6
- - name: image
7
- dtype: string
8
- - name: target
9
- dtype: string
10
- - name: instrument
11
- dtype: string
12
- - name: filter
13
- dtype: string
14
- - name: date_obs
15
- dtype: string
16
- - name: exptime
17
- dtype: 'null'
18
- - name: ra
19
- dtype: 'null'
20
- - name: dec
21
- dtype: 'null'
22
- - name: program
23
- dtype: string
24
- - name: image_path
25
- dtype: string
26
- - name: width
27
- dtype: int64
28
- - name: height
29
- dtype: int64
30
- - name: total_pixels
31
- dtype: int64
32
- - name: mean_intensity
33
- dtype: float64
34
- - name: std_intensity
35
- dtype: float64
36
- - name: min_intensity
37
- dtype: float64
38
- - name: max_intensity
39
- dtype: float64
40
- - name: median_intensity
41
- dtype: float64
42
- - name: skewness
43
- dtype: float64
44
- - name: kurtosis
45
- dtype: float64
46
- - name: dynamic_range
47
- dtype: float64
48
- - name: noise_level
49
- dtype: float64
50
- - name: noise_std
51
- dtype: float64
52
- - name: signal_to_noise
53
- dtype: float64
54
- - name: saturated_pixels
55
- dtype: int64
56
- - name: saturation_percentage
57
- dtype: float64
58
- - name: is_saturated
59
- dtype: bool
60
- - name: cosmic_rays
61
- list:
62
- list: int64
63
- - name: hot_pixels
64
- list:
65
- list: int64
66
- - name: bad_pixels
67
- list: 'null'
68
- - name: artifact_count
69
- dtype: int64
70
- - name: quality_score
71
- dtype: float64
72
- splits:
73
- - name: train
74
- num_bytes: 18984918
75
- num_examples: 2709
76
- download_size: 3821834
77
- dataset_size: 18984918
78
- configs:
79
- - config_name: default
80
- data_files:
81
- - split: train
82
- path: data/train-*
83
- ---
 
 
 
 
 
 
 
1
+ # JWST Quality Analysis Dataset
2
+
3
+ ## Overview
4
+
5
+ This dataset contains comprehensive quality analysis for 2,709 JWST (James Webb Space Telescope) NIRCam images from the MAST archive. Each image has been automatically analyzed for quality metrics, artifact detection, and noise characteristics.
6
+
7
+ ## Dataset Information
8
+
9
+ - **Size**: 2,709 images
10
+ - **Format**: JSONL (JSON Lines)
11
+ - **Source**: JWST NIRCam observations from MAST
12
+ - **Targets**: M16, NGC 3132, NGC 3324, SMACS 0723, Stephan's Quintet
13
+
14
+ ## Quality Metrics
15
+
16
+ Each image includes:
17
+
18
+ ### Basic Statistics
19
+ - `mean_intensity`, `std_intensity`, `min_intensity`, `max_intensity`
20
+ - `median_intensity`, `skewness`, `kurtosis`, `dynamic_range`
21
+
22
+ ### Noise Analysis
23
+ - `noise_level`, `noise_std`, `signal_to_noise`
24
+
25
+ ### Saturation Analysis
26
+ - `saturated_pixels`, `saturation_percentage`, `is_saturated`
27
+
28
+ ### Artifact Detection
29
+ - `cosmic_rays`: List of cosmic ray locations [x, y, area]
30
+ - `hot_pixels`: List of hot pixel locations [x, y, area]
31
+ - `bad_pixels`: List of bad pixel locations [x, y, area]
32
+ - `artifact_count`: Total number of artifacts
33
+
34
+ ### Quality Assessment
35
+ - `quality_score`: Overall quality score (1-10 scale)
36
+
37
+ ## Use Cases
38
+
39
+ ### For Researchers
40
+ - **Quality Screening**: Filter images by quality score for analysis
41
+ - **Artifact Cataloging**: Identify and locate artifacts for cleaning
42
+ - **Statistical Analysis**: Study image quality across different targets/filters
43
+ - **Quality Benchmarking**: Compare quality across different observations
44
+
45
+ ### For Machine Learning
46
+ - **Training Data**: Train quality assessment models
47
+ - **Feature Engineering**: Use quality metrics as features
48
+ - **Validation**: Quality scores for model evaluation
49
+
50
+ ## Methodology
51
+
52
+ The quality analysis was performed using:
53
+ - **OpenCV** for image processing and artifact detection
54
+ - **NumPy/SciPy** for statistical analysis
55
+ - **Parallel processing** for efficient analysis of large datasets
56
+
57
+ Quality scores are calculated based on:
58
+ - Signal-to-noise ratios
59
+ - Saturation levels
60
+ - Artifact counts
61
+ - Dynamic range
62
+
63
+ ## Dataset Statistics
64
+
65
+ - **Mean Quality Score**: 8.39/10
66
+ - **Quality Score Range**: 6.0 - 10.0
67
+ - **Images with Artifacts**: All images contain some artifacts (typical for astronomical data)
68
+ - **Saturated Images**: 0 (no significant saturation detected)
69
+
70
+ ## Citation
71
+
72
+ If you use this dataset in your research, please cite:
73
+
74
+ ```bibtex
75
+ @dataset{jwst_quality_analysis_2024,
76
+ title={JWST Quality Analysis Dataset},
77
+ author={Your Name},
78
+ year={2024},
79
+ url={https://huggingface.co/datasets/norbertm/jwst-quality-analysis-dataset}
80
+ }
81
+ ```
82
+
83
+ ## License
84
+
85
+ This dataset is provided for research purposes. Please refer to the original JWST data usage policies from MAST.
86
+
87
+ ## Contact
88
+
89
+ For questions or feedback about this dataset, please open an issue on the Hugging Face repository.