Neural network based SOM for multispectral image segmentation in RGB and HSV color space

Abstract

Segmentation is the process of partitioning an image into number of meaningful images as segments or clusters. The segmentation is initial but important process which is used to locate boundaries and objects in images. This paper is concerned with segmentation of color satellite images using neural network based kohonen's self-organizing maps. This unsupervised competitive network is used to visualize and interpret large data sets. In this paper, test images are segmented in RGB and HSV color space using self-organizing map and the segmentation results are compared using error image, peak signal to noise ratio, and execution time. The efficiency of proposed method is tested with Landsat and Terra (MODIS sensor) satellite images.

Accurate and Efficient Segmentation of Optic Disc and Optic Cup in Retinal Images Integrating Multi-View Information

“…So, the multi-feature spaces need to be integrated to complement each other's advantages for adequately describing the OD. Considering that the red color plane has an unobvious contrast for blood vessel and gives a better contrast of the OD region [17], and the HSV color space can easily separate the intensity information from the color information and retrieve more information [34]. Then, the extended LSACM uses a multi-dimensional feature space (d = 5) where the individual vector element is taken from red color plane, vessel-free red color plane, and each channel from vessel-free HSV color space, to represent an image point x.…”

“…So, the multi-feature spaces need to be integrated to complement each other's advantages for adequately describing the OC. Considering that the intensity information from the color information can be easily separated and more information can be retrieved in the HSV color space [34], and the green color plane provides a better contrast for OC [37]. Then, we use a multi-dimensional feature space (m = 5) which consists of red color plane, vessel-free green color plane and each channel from vessel-free HSV color space to describe an image point x.…”

Image Segmentation Using Fuzzy Inference System on YCbCr Color Model

“…A segmentation system based on neural networks has been proposed by Ganesan et al [5], they performed an unsupervised competitive neural network based in Kohonen's self-organizing maps (SOM). They used color satellite images.…”