A Fast Multi-Object Extraction Algorithm Based on Cell-Based Connected Components Labeling

Abstract

Qingyi GU†a) , Takeshi TAKAKI †b) , Nonmembers, and Idaku ISHII †c) , Member SUMMARYWe describe a cell-based connected component labeling algorithm to calculate the 0th and 1st moment features as the attributes for labeled regions. These can be used to indicate their sizes and positions for multi-object extraction. Based on the additivity in moment features, the cell-based labeling algorithm can label divided cells of a certain size in an image by scanning the image only once to obtain the moment features of the labeled regions with remarkably reduced computational complexity and memory consumption for labeling. Our algorithm is a simple-onetime-scan cell-based labeling algorithm, which is suitable for hardware and parallel implementation. We also compared it with conventional labeling algorithms. The experimental results showed that our algorithm is faster than conventional raster-scan labeling algorithms. key words: fast multi-object extraction, connected component labeling

Fast Tracking System for Multi-colored Pie-shaped Markers

“…To emulate a previous evaluation of the cell-based labeling algorithm (Gu et al 2012b), we assessed the execution times and memory consumption of our multicolored pie-shaped marker tracking algorithm using a PC by calculating the moment features for each labeled colored sector.…”

“…After converting source images I(r) into binary images B(r), a cell-based labeling algorithm (Gu et al 2012b) can be used to determine the moments of labeled…”

“…In our algorithm, there was a trade-off between robustness with small and narrow connected components in an image and the accuracy of the moment feature calculation, which was affected by the threshold that determined the active cell. This trade-off in cell-based labeling was evaluated in a previous study (Gu et al 2012b). …”

“…To overcome this restriction, we use a cell-based labeling algorithm (Gu et al 2012b) to extract the multi-object size and position at a high frame rate by implementing the algorithm on FPGA. The cell-based labeling algorithm requires clearly binarized images, and its acceleration is essential for real-time feedback control in well-designed, artificially illuminated factory conditions where high-speed operations are executed.…”

Automatic extraction of road segments from high resolution satellite images

“…In the output image, the pixel value in each position has been replaced by the mean intensity value of the pixels belonging to the same region. Individual segments can now be extracted using the connected component technique [14]. 8-neighborhood has been considered for this case.…”