Mobiusi/Experiment-Design-Sketch-Image-Classification-Dataset

Experiment Design Sketch Image Classification Dataset

In the field of industrial manufacturing, the design process often relies on a large number of design sketches that need to be quickly converted into actual engineering designs during the subsequent manufacturing stages. However, manually processing these sketches is often time-consuming and prone to errors, currently relying mainly on manual labeling and conversion by designers, which is inefficient and unstable. Existing automation solutions are mostly limited to simple graphic recognition and difficult to accurately handle complex design sketches. The Experiment Design Sketch Image Classification Dataset aims to improve automated recognition techniques by using high-quality, large-scale annotated sketch images to help solve key challenges in sketch recognition and classification. Data collection involves scanning design sketches using high-resolution scanners to ensure image quality under conditions of no reflection and uniform light source. Quality control measures include multiple rounds of annotation verification and consistency checks, and expert review to ensure the precision and consistency of annotations. The annotation team consists of 30 professionals with design backgrounds. Data preprocessing includes image denoising, contour enhancement, and size normalization, stored in JPG format, organized into categories using a directory structure. The annotation accuracy of this dataset exceeds 98%, maintaining a high degree of consistency and integrity. By introducing semi-automated annotation tools and data augmentation algorithms, the accuracy and robustness in classification tasks are enhanced. Experiments show that the recognition system applying this dataset can reduce sketch processing time by more than 30%, improving the efficiency of design and manufacturing. Compared to similar datasets, this dataset offers higher annotation accuracy and a more diverse range of sketch types, covering multiple design scenarios with strong scalability and generality.

Technical Specifications

Field	Type	Description
file_name	string	File name
quality	string	Resolution
design_type	string	The type of design of the sketch, e.g., mechanical, electronic, architectural, etc.
sketch_complexity	integer	The complexity degree of the sketch design, usually indicated by the number of details or element density.
drawing_tool	string	The type of drawing tool used to create the sketch, such as a pen, pencil, or CAD software.
color_palette	string	The color scheme used in the design sketch, such as monochrome, two-tone, or multicolor.
symmetry	boolean	Indicates whether the sketch possesses symmetry.
annotation_presence	boolean	Indicates the presence of text or symbolic annotations on the sketch.

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