did=312 task=did=312 YACVID - University of Leon - Edge profile milling head tool data set - Details

Yet Another Computer Vision Index To Datasets (YACVID) - Details

Stand: 2024-03-19 11:29:30 - Overview

ok - did=312 - task=task - dviews=1362
Attribute Current content New content
Name (Institute + Shorttitle)University of Leon - Edge profile milling head tool data set 
Description (include details on usage, files and paper references)This data set comprises 144 images of an edge profile cutting head of a milling machine. The head tool contains a total of 30 cutting inserts. The cutting head is formed by 6 diagonals of inserts in radial direction along the tool perimeter, encompassing 5 inserts per diagonal in axial direction. Positions of the last and first inserts of consecutive diagonals are aligned in the same vertical. Therefore, even though we have 30 inserts, there are 24 equally spaced positions of inserts along the tool perimeter. Additionally, inserts are squared shape with four 90º indexable cutting edges. Inserts are fastened with a screw. Rake angle is 0.

Images were taken with a monochrome camera Genie M1280 1/3’’ with active resolution of 1280 × 960 pixels. We used AZURE-2514MM fixed lens with 25 mm focal length and resolution of 2 mega-pixel. We used two compact bar shape structures with high intensity LED arrays BDBL-R(IR)82/16H.

Once a set of 30 inserts are placed in the cutting head, we can start taking images. We take an image each 15º, when a new insert is aligned with the camera. In this way, a total of 24 images are taken and the same insert is captured in different positions. Then we replace the 30 inserts with different ones and the process is repeated. After the sixth iteration, we have created our data set of 24 × 6 = 144 images. The number of inserts per image varies from 8 to 10. In total 180 unique inserts were used.

The data set comprises six parts:
1. The 144 images of the edge milling head. (Only complete inserts have been considered for creating the ground truth, discarding partly visible ones.)
2. A list of coordinates for each central point of the screw that fastens an insert.
3. A mask for each image is created, with black background and white circles of 40 pixels centred at the previous coordinates.
4. A list of the coordinates of the end points of the ideal cutting edges.
5. A set of ground truth images with black background and white 1-pixel-wide lines indicating the ideal cutting edges.
6. A list of broken (19 inserts) and unbroken inserts (161 inserts).

Related publication:
L. Fernández-Robles, G. Azzopardi, E. Alegre, and N. Petkov, Cutting edge localisation in an edge profile milling head, Computer Analysis of Images and Patterns - 16th International Conference, CAIP 2015, Valletta, Malta, September 2-4, Proceedings, Part II, Vol. 9257, pp. 336-347, 2015.
 
URL Link http://pitia.unileon.es/varp/node/395 
Files (#)
References (SKIPPED)
Category (SKIPPED) 
Tags (single words, spaced)milling head tool inserts localization object cutting tool edge profile tool wear monitoring 
Last Changed2024-03-19 
Turing (2.12+3.25=?) :-)