Dataset Viewer
Auto-converted to Parquet Duplicate
file_name
stringclasses
4 values
quality
stringclasses
4 values
defect_type
stringclasses
4 values
defect_location
stringclasses
4 values
severity_level
stringclasses
3 values
bounding_box_coordinates
stringclasses
4 values
surface_condition
stringclasses
4 values
light_condition
stringclasses
1 value
image_quality
stringclasses
2 values
8422f40db2daec16658744c0586b7185.jpg
1080*1414
Cracks and wear
Slits and openings in the central part of the mold
Moderate
(150, 100, 300, 200)
Slight surface wear with metal shavings
Natural light
Clear
a4e76f92f1ba140093566059c85336ed.jpg
1080*1403
Crack
Multiple slits in the central part of the mold
Medium
(100, 100, 200, 200)
Slightly worn
Natural light
Clear
ddfdfae294b72f292dd207659a87d6a8.jpg
1074*1888
Notch
Edge of the circular hole
Moderate
(100, 150, 200, 250)
Relatively rough, with wear
Natural light
Relatively clear
f4d4b8eee780f6c263646b3d7fe17f3d.jpg
1920*1038
No significant defect observed
None
None
None
Good
Natural light
Clear

Mold Template Defect Recognition Dataset

In the industrial sector, mold quality inspection is crucial for ensuring product integrity but faces challenges such as inconsistent defect detection and high rates of false negatives. Existing solutions often rely on manual inspection, which is time-consuming and prone to human error, leading to inefficiencies. This dataset aims to address these challenges by providing a comprehensive collection of labeled images that enhance machine learning models for accurate defect recognition in mold templates. Data was collected using high-resolution cameras in controlled industrial environments, ensuring clarity and consistency. Quality control measures included multiple rounds of labeling, consistency checks among annotators, and expert reviews to maintain high accuracy. The images are stored in JPG format, organized in directories by defect type, facilitating easy access and processing.

Technical Specifications

Field Type Description
file_name string File name
quality string Resolution
defect_type string The types of defects appearing on the mold, such as cracks and pores.
defect_location string A detailed description of the defect's specific location on the mold.
severity_level integer The severity level of the defect as defined by the standards.
bounding_box_coordinates string Coordinates of the bounding box that marks the defect in the target detection task.
surface_condition string The overall condition of the mold surface, including the presence of oxidation or wear.
light_condition string The lighting conditions when capturing the image, such as natural light or artificial light.
image_quality string The quality assessment of the image, such as clarity or blurriness.

Compliance Statement

Authorization Type CC-BY-NC-SA 4.0 (Attribution–NonCommercial–ShareAlike)
Commercial Use Requires exclusive subscription or authorization contract (monthly or per-invocation charging)
Privacy and Anonymization No PII, no real company names, simulated scenarios follow industry standards
Compliance System Compliant with China's Data Security Law / EU GDPR / supports enterprise data access logs

Source & Contact

If you need more dataset details, please visit Mobiusi. or contact us via contact@mobiusi.com

Downloads last month
9