A regression modelling approach for optimizing segmentation scale parameters to extract buildings of different sizes

Jozdani, Shahab; Momeni, Mehdi; Johnson, Brian Alan; Sattari, Mohammad

doi:10.1080/01431161.2017.1390273

Cited by 18 publications

(15 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As a result, to extract different land cover classes, it is more reasonable to use multiple SPs, each of which is appropriate for a separate land cover class. Such a multi-scale/level approach can positively affect the quality of extraction of different land cover classes, and thus can improve the accuracy of final classification results, as reported in several studies [17][18][19][20].…”

mentioning

confidence: 96%

“…In a recent study, Jozdani et al [20] developed a scene-independent unsupervised approach to optimizing the SP to extract urban buildings of different sizes. In that research, by assuming that the sizes of buildings in a given urban block are close to each other, a degree-2 polynomial regression model was established that associated appropriate SPs with the median size of buildings and the spatial resolution of the image.…”

Section: Related Workmentioning

confidence: 99%

“…Studies on parametrizing SP can be categorized into two general groups: supervised approaches [21,30,31,40], and unsupervised approaches [17,20,22,29]. With supervised methods, the goal is to identify an optimal segmentation (or set of segmentations) by evaluating the overlap of ground-truth reference polygons and computer-generated image segments.…”

Section: Related Workmentioning

confidence: 99%

“…In practice, multiple segmentation parameters are usually tested and compared, and "suitable" segmentation parameters (i.e., those that do not cause excessive over-or undersegmentation) are determined either manually (by trial and error [25][26][27]) or automatically (e.g., using supervised or unsupervised parameter selection algorithms [17,20,22,[28][29][30][31]). Although manual parameter selection through visual assessment could lead to the most optimal segmentation results if done by a skilled image analyst, the process is typically time consuming and/or computationally intensive.…”

Section: Segmentation Parameter Selectionmentioning

confidence: 99%

“…For this purpose, they first performed multiple segmentations with different SP values (i.e., 10,15,20,25,30,35,40,45,50,55,60, 70, 80, 90, and 100) and then calculated several features (i.e., spectral, textural, geometric, etc.) for the resulting image objects at different scale levels.…”

Section: Related Workmentioning

confidence: 99%

See 4 more Smart Citations

Identifying Generalizable Image Segmentation Parameters for Urban Land Cover Mapping through Meta-Analysis and Regression Tree Modeling

Johnson

Jozdani

2018

Remote Sensing

Self Cite

View full text Add to dashboard Cite

Abstract:The advent of very high resolution (VHR) satellite imagery and the development of Geographic Object-Based Image Analysis (GEOBIA) have led to many new opportunities for fine-scale land cover mapping, especially in urban areas. Image segmentation is an important step in the GEOBIA framework, so great time/effort is often spent to ensure that computer-generated image segments closely match real-world objects of interest. In the remote sensing community, segmentation is frequently performed using the multiresolution segmentation (MRS) algorithm, which is tuned through three user-defined parameters (the scale, shape/color, and compactness/smoothness parameters). The scale parameter (SP) is the most important parameter and governs the average size of generated image segments. Existing automatic methods to determine suitable SPs for segmentation are scene-specific and often computationally intensive, so an approach to estimating appropriate SPs that is generalizable (i.e., not scene-specific) could speed up the GEOBIA workflow considerably. In this study, we attempted to identify generalizable SPs for five common urban land cover types (buildings, vegetation, roads, bare soil, and water) through meta-analysis and nonlinear regression tree (RT) modeling. First, we performed a literature search of recent studies that employed GEOBIA for urban land cover mapping and extracted the MRS parameters used, the image properties (i.e., spatial and radiometric resolutions), and the land cover classes mapped. Using this data extracted from the literature, we constructed RT models for each land cover class to predict suitable SP values based on the: image spatial resolution, image radiometric resolution, shape/color parameter, and compactness/smoothness parameter. Based on a visual and quantitative analysis of results, we found that for all land cover classes except water, relatively accurate SPs could be identified using our RT modeling results. The main advantage of our approach over existing SP selection approaches is that our RT model results are not scene-specific, so they can be used to quickly identify suitable SPs in other VHR images.

show abstract

mentioning

confidence: 96%

Section: Related Workmentioning

confidence: 99%