A desktop version is available with more powerful, stable features.
Open desktop version
Set scoring and filter rules first, then import images, start checking, and download results. Batch files show only the first 20 previews; all images will be processed.
Set Scoring Rules
Overall score is based on sharpness, exposure, contrast, noise, and compression artifacts. Adjust thresholds and weights by use case.
Set before processing
Basic parametersFiner settings take longer for large images
Images pass when the overall score reaches this value and size rules are met.
Longest side used for checking. Original files are not changed.
Choosing a common use case adjusts metric weights automatically.
Keeping original format preserves animated files; conversion depends on current browser support.
Enter 0 for no width limit.
Enter 0 for no height limit.
Custom WeightsNormalized by current ratio during detection
Current weight total: 100%
Select images to check
Supports common images such as JPG, PNG, WebP, GIF, and BMP. When selecting multiple files, only the first 20 are shown; detection and packaging process all files.
No image selected
Click to choose images, or drag them here
Images are not uploaded to the server. Detection is done in this page.
Image Preview
Results and download
Results are written to a ZIP by passed, failed, and processing failed, with a CSV score report. Filenames use the original names by default.
Waiting to scan
Total images
0
Passed
0
Failed
0
Average score
0
Please select images before starting detection.
Preview
File Name
Size
Clarity
Exposure
Contrast
Noise
Compressed
Overall
Category
Tip
No results generated yet.
Instructions
Software Usage Instructions
Set detection parameters: First configure the passing score, analysis size, scoring preset, output format, and minimum size. The passing score is used to determine whether an image is placed in the qualified category.
Select a scoring preset:
Comprehensive quality check: Suitable for everyday image quality screening.
Photo culling: Focuses more on sharpness and focus performance.
E-commerce images: Focuses more on exposure, sharpness, and image stability.
Document scanning: Focuses more on sharpness and contrast.
Custom weights: Allows manual adjustment of the scoring proportions for sharpness, exposure, contrast, noise, and compression artifacts.
Import image files: Click to select images or drag images to the upload area. Common formats such as JPG, PNG, WebP, GIF, and BMP are supported. When selecting in batches, the interface only shows previews of the first 20 images, but the actual detection will process all images.
Start detection: After entering the detection and download step, click “Start Detection”. The tool will analyze sharpness, exposure, contrast, noise, and compression artifacts one by one, and generate an overall score.
View results: After detection is complete, you can view the total number, qualified count, unqualified count, average score, and the first 20 scoring details. Each image will display its dimensions, scores for each metric, overall score, and prompt information.
Download results: Click “Download ZIP Results” to download the categorized images and scoring details; click “Export CSV” to download the score sheet separately for later organization and filtering.
FAQ
A: Sharpness is used to determine whether edge details are crisp; exposure is used to determine whether the image is too bright or too dark; contrast is used to assess light and dark gradations; noise is used to assess grain and speckles; compression artifacts are used to detect blocky traces caused by compression such as JPEG.
A: The default score of 65 is suitable for general screening. For product images, asset libraries, or photo culling, you can raise the threshold according to actual requirements; if you only want to exclude images that are clearly blurry, too dark, or have high noise, you can lower the threshold appropriately.
A: No. The analysis size only affects the temporary image size used during detection and does not modify the original image. The larger the size, the more thoroughly details can be assessed, but processing will be slower.
A: The ZIP contains categorized folders for “Passed,” “Failed,” and “Processing Failed,” along with a score details CSV and detection settings JSON. Image file names will retain the original file names as much as possible, and if duplicates occur, numbers will be automatically appended to distinguish them.
A: The overall score considers not only sharpness, but also exposure, contrast, noise, compression artifacts, and minimum size requirements. If an image is sharp but too dark, has high noise, obvious compression artifacts, or insufficient dimensions, it may also be classified as failed.
A: You can set the analysis size to 800px or 1200px and process images in batches. Ultra-high-resolution images, large quantities, or unified format conversion will consume more computing resources.
Related Tools
Batch Image Top/Bottom Text Tool
Add custom text to the top or bottom of images in bulk, w...
