Scale Object Selection (SOS) through a hierarchical segmentation by a multi-spectral per-pixel classification

Abstract

International audienceIn high resolution multispectral optical data, the spatial detail of the images are generally smaller than the dimensions of objects, and often the spectral signature of pixels is not directly representative of classes we are interested in. Thus, taking into account the relations between groups of pixels becomes increasingly important, making object­oriented approaches preferable. In this work several scales of detail within an image are considered through a hierarchical segmentation approach, while the spectral information content of each pixel is accounted for by a per­pixel classification. The selection of the most suitable spatial scale for each class is obtained by merging the hierarchical segmentation and the per­pixel classification through the Scale Object Selection (SOS) algorithm. The SOS algorithm starts processing data from the highest level of the hierarchical segmentation, which has the least amount of spatial detail, down to the last segmentation map. At each segmentation level, objects are assigned to a specific class whenever the percentage of pixels belonging to the latter, according to a pixel­based procedure, exceeds a predefined threshold, thereby automatically selecting the most appropriate spatial scale for the classification of each object. We apply our method to multispectral, panchromatic and pan­sharpened QuickBird images

Technical note: Mapping surface-saturation dynamics with thermal infrared imagery

“…Silasari et al (2017) applied an automatic image classification for unimodal distributions based on a threshold pa-rameter that needs to be calibrated to specific image conditions (in this case, the brightness of VIS images).This is only straightforward in cases where the temperature distribution between water and the surrounding environment is clearly bimodal. Chini et al (2017) presented a parametric adaptive thresholding algorithm especially suited for images that do not show a clear bimodal distribution. The algorithm makes use of an automatic selection of image subsections with clear bimodal distributions, a hierarchical split-based approach (HSBA), and a subsequent parameterization of the distributions of the two pixel classes.…”

“…The algorithm makes use of an automatic selection of image subsections with clear bimodal distributions, a hierarchical split-based approach (HSBA), and a subsequent parameterization of the distributions of the two pixel classes. Since the two decomposed distributions might still overlap to a certain extent, Chini et al (2017) advise complementing the decomposed distribution information with contextual information of the image for the final generation of a binary image, instead of selecting a single threshold value between the two decomposed distributions. Several approaches are available in the literature for including contextual information in the classification of a single spectral image, such as mathematical morphology (Chini et al, 2009) or second-order textural parameters (Pacifici et al, 2009).…”

“…Several approaches are available in the literature for including contextual information in the classification of a single spectral image, such as mathematical morphology (Chini et al, 2009) or second-order textural parameters (Pacifici et al, 2009). Chini et al (2017) suggested a region-growing algorithm where the seeds and the stopping criteria are constrained by the identified distribution of the class of interest (here, saturation). Figure 7.…”

Towards a 20 m Global Building Map from Sentinel-1 SAR Data

“…Synthetic aperture radars (SARs) are rarely the primary source of information for land cover classification, especially when multispectral optical images at high spatial resolution are available [8][9][10][11][12]. SAR data are more often used merely as a complementary or alternative data source in case of unfavorable atmospheric conditions or to identify classes that have a highly distinctive scattering behavior.…”

Spatial Up-Scaling Correction for Leaf Area Index Based on the Fractal Theory

“…Scale effect are highly correlated with many issues, such as parameters inversion [1,2], objects classification [3,4], agricultural and ecological data assimilation [5,6], remote sensing product validation [7,8], and data fusion [9]. Therefore, it is necessary to comprehensively explore the scaling effect correction.…”