Examining the object-based and pixel-based image analyses for developing stand volume estimator model

Abstract

In the last two decades there has been significant leap on the spatial resolution of the satellite digital images which may be very useful for estimating stand parameter required for forest as well as environment management. This paper describes development of stand volume estimator models using SPOT 6 panchromatic and multispectral images with an object-based digital image analysis (OBIA) and conventional pixel-based approaches. The data used include panchromatic band with1.5m spatial resolution, and multispectral band with6m spatial resolution. The proposed OBIA technique with mean-shift algorithm was functioned to derive a canopy cover variable from the fusion of the panchromatic and multispectral, while the pixel-based vegetation index was used to develop model with an original pixel-size of 6 m. The estimator models were established based on 65 sample plots both measured in the field and images. The study found that the OBIA provides more accurate identification with Kappa Accuracy (KA) of 71% and Overall Accuracy (OA) of 86%. The study concluded that the best stand volume estimation model is the model that developed from the canopy cover (C) derived from OBIA i.e., v = 13.47e<sup>0.032C</sup> with mean deviation of only 0.92%, better than the model derived from conventional pixel-based approach, i.e., v = 0.0000067e<sup>16.48TNDVI</sup> with a mean deviation of 5.37%.

Crown closure segmentation on wetland lowland forest using the mean shift algorithm

“…It is shown that the accuracies do not vary at all between different hs values, either within the same images, or different images, for both the Pleiades-1B and SPOT-6 images. These findings are in line with the findings of some previous studies [4], [14], [17], where the variation in hs parameter values did not change the classification accuracy of segmentation significantly.…”

“…As shown in Figures 6 and 7, the greater the hr, the less accurate the segmentation. In line with previous studies, the variation in the hr values causes the segmentation accuracy to be low [4], [7], [15], [26], [27]. In this study, the accurate of segmentation of crown closure using Pleiades-1B were provided by hr values between 7 and 42 Figure 6; while for the SPOT-6, the accurate segmentation were provided by hr values ranging from 4 to 18 (Figure 7).…”

“…Inventory of wetland lowland forest ecosystems is quite challenging and not efficient, if all done by terrestrial methods. Therefore, a combination of methods between terrestrial surveying and remote sensing techniques utilizes The main challenge in using high-resolution and very-high resolution imageries with visual method were the consistency and accuracy on measuring stand dimensions such as measuring the percentage of crown closure, number of trees and crown diameter [4]. To date, there is no practical algorithm available for measuring the percentage of crown cover automatically.…”

“…The mean shift (MS) algorithm is an object-based segmentation algorithm that consists of three parameters, namely: spatial radius (hs), range radius (hr), and minimum region size (M) [14]. Several studies have used the MS algorithm in identifying land use in the optimal GeoEye-1 composite image parameter combination (hs: 30, hr: 35, M: 85.2) [7], the segmentation studies detect forest changes due to storms on Formosat-2 red band images with optimal parameter combinations (hs: 17, hr:2) [15], the segmentation studies of pine, deciduous, and spruce trees on light detection and ranging (LiDAR) with optimal parameter combination (hs: 50, hr: 50, M: 200) [16], the segmentation study of crown closure on the Nipah ecosystem used composite images of RGB unmanned aerial vehicle (UAV) with optimal combination parameters (hs: 10, hr: 10, M: 50) [17], the segmentation studies of crown closure for the development of a stand prediction model in the SPOT-6 RGB composite image for optimal segmentation combination (hs: 5, hr: 21, M: 4) [4]. The determination of this parameter is very important because it affects the segmentation results [5], [18], [19].…”

Evaluation of weighted fusion for scalar images in multi-sensor network

“…Feature method [11] fuses the characteristic of multiple sources. Pixel method [12] is fundamental and commonly used. It fuses pixels (scalar) in the sources and preserves most of the original details.…”