Boundary Displacement Error

The use of visible color difference in the quantitative evaluation of color image segmentation, Hsin-Chia Chen and Sheng-Jyh Wang

Performance Measures for Video Object Segmentation and Tracking Erdem, Sankur, Tekalp

struct FBoundary

Learning to Detect Natural Image Boundaries Using Local Brightness, Color, and Texture Cues David R. Martin, Member, IEEE, Charless C. Fowlkes, and Jitendra Malik, Member, IEEE

Members

• dmax::Float64

FMeasure

FMeasure Regions

Quantitative evaluation of color image segmentation results M. Borsotti a, P. Campadelli a,2, R. Schettini b,

Multiresolution Color Image Segmentation Jianqing Liu and Yee-Hong Yang, Senior Member, IEEE

http://www.cv-foundation.org/openaccess/contentcvpr2013/papers/Pont-TusetMeasuresandMeta-Measures2013CVPRpaper.pdf

Measures and Meta-Measures for the Supervised Evaluation of Image Segmentation Jordi Pont-Tuset and Ferran Marques. Universitat Politecnica de Catalunya BarcelonaTech

Precision

Quantitative evaluation of color image segmentation results M. Borsotti a, P. Campadelli a,2, R. Schettini b,

where $R$ is the number of regions in the segmented image, $A_i$ is the area, or the number of pixels of the ith region $i$, and $e_i$ the color error of region $i$. e is defined as the sum of the Euclidean distance of the color vectors between the original image and the segmented image of each pixel in the region, while $R(A_i)$ represents the number of regions having an area equal to $A_i$.

RandIndex

Segmentation Covering

An Entropy-based Objective Evaluation Method for Image Segmentation Hui Zhang*, Jason E. Fritts and Sally A. Goldman

Given an image I of ($n \times m$) , $S_I = nm$

$ Hl(I) = - \sum\limits{j=1}^N \dfrac{Sj}{SI} log(\dfrac{Sj}{SI}) $

Variation of Information

Zéboudj, Rachid. Filtrage, seuillage automatique, contraste et contours: du pré-traitement à l'analyse d'image. Diss. Saint-Etienne, 1988. Unsupervised Evaluation of Image Segmentation Application to Multi-spectral Images

"This contrast takes into account the internal and external contrast of the regions measured in the neighborhood of each pixel"

"

function boundary_map(seg::Matrix{T}) where T<:Integer

From a segmentation, compute a binary boundary map with 1 pixel wide boundaries. The boundary pixels are offset by 1/2 pixel towards the origin from the actual segment boundary.

function LineNormals2D(vertices)

This function calculates the normals, of the line points using the neighbouring points of each contour point, and forward an backward differences on the end points

´´´julia function colorerrorsum(image::Matrix, segments::Matrix{Integer) ´´´ Computes $\sum\limits_{i=1}^R \dfrac{e_i^2}{\sqrt{A_i}}$ where $R$ is the number of regions in the segmented image, $A_i$ is the area, or the number of pixels of the ith region $i$, and $e_i$ the color error of region $i$. e is defined as the sum of the Euclidean distance of the color vectors between the original image and the segmented image of each pixel in the region.

"

function relabel(c1::Matrix}, part_bimap=Dict{Integer,Integer}()) where T<:Integer

It relabels a partition in scanning order. Bimaps are the look up tables of the relabeling.

• author Jordi Pont Tuset <jordi.pont@upc.edu>