Star identification based on euclidean distance transform, voronoi tessellation, and k-nearest neighbor classification

DIRT/3D: 3D root phenotyping for field grown maize (Zea mays)

“…The Euclidean Distance Transform (EDT) of the image allows for direct detection of the local minima (Fabbri, Costa et al 2008). In that way, watersheds assign each pixel to a unique component and allows the distinction entangled roots per level set image (Roshanian, Yazdani et al 2016).…”

Improved Grid Algorithm Based on Star Pair Pattern and Two-dimensional Angular Distances for Full-Sky Star Identification

“…In recent years, some new star identification algorithms were developed. For instances, the shortest distance transform based algorithm [17], Euclidean distance transform based algorithm [18], and the multi-poles algorithm [19] utilize the features composed of multiple vertices so that the matching process is more complicated. Although the robustness is high, the identification time is long.…”

A reliable and fast lost-in-space mode star tracker

“…In general, the intensity of the star (I ) in the image is inversely proportional to its relative apparent (M v ) in the star catalog, which can be deduced from Eq 4. 47.…”

“…The existing approaches such as Pyramid [26] , k -nearest neighbors [47] rely on the nearest four stars for star pattern recognition. Thus, we select four stars as a worst case of the number of nearest stars missed due to magnitude uncertainty.…”

“…G. Zhang proposed techniques[45,46] which utilized cyclic and radial features for comparison. A star identification based on Voronoi tessellation was proposed recently in[47]. Other SPR techniques which fall under this category are[48][49][50][51][52][53][54][55].…”