基于颜色和纹理的森林影像分割方法

提出一种基于影像颜色散度和纹理信息的森林影像分割方法.对输入影像进行颜色量化,然后利用量化得到的索引矩阵计算各像素的颜色散度;基于多分辨率策略进行纹理分析,通过区域生长形成初始分割区域;统计各区域Laws纹理能量,合并过分割区域.实验表明,森林影像的分割结果与人的主观视觉感知具有良好的一致性,识别出的地面和树木可为匹配和三维建模提供可靠依据.     

Growing stock estimation through stratified sampling on satellite remote sensing data

Aerial photographs are being extensively used for forest surveys i.e. forest cover type mapping, assessment of growing stock, estimation of area, vegetation studies, etc. Satellite remote sensing technology offers new possibilities and scope for achieving some of the above applications with higher accuracy and reliability. The assessment of growing stock through subjective stratification has also become possible with increase in spatial and spectral resolution. In the present study, the LANDSAT TM FCC (1966) has been visually interpreted on the basis of tonal characteristics. Stratified random sampling method has then used for determination of number of sample units for collection of ground Inventory data. Using regression equations, the volume per hectare of individual forest cover types were calculated. Satellite remote sensing data has been used for initial stratification, distribution of sample plots and calculation of area under various forest cover types. Estimate has been made for available commecial and non-commercial growing stocks in the study area.     

A novel method for enhancement of radiometric resolution using image fusion

The Resourcesat-I satellite is equipped with different types of sensors with varied characteristics. For the effective utilization of the available multi-sensor, multi-temporal, multi-spectral and multi-radiometric data from these sensors, fusion of digital image data has become a valuable technique. Image fusion enhances the information content and helps in better discrimination of various land cover types. The Resourcesat-1 has equipped with three sensors, AWiFS, LISS-III and LISS-IV, which are having identical spectral resolutions, with different spatial, radiometric and temporal resolutions. The spatial resolutions ratio of the data set for merging are required to be maximum of 1:6, where as the data sets (AWiFS and LISS-III) that are used in the current study are having the ratio of 1:2.5 approximately. A novel merging technique is designed, which retains the multi-spectral response of the input data in the output data. The merged data set provides the higher spatial and radiometric resolutions. In order to evaluate the fusion merits quantitatively, all the data sets are digitally classified and studied the output classes for homogeneity and clear discrimination. A comprehensive comparative study is carried out between the fused image and the LISS-III image based on the contingency matrix and the scatter plots, which demonstrates the strength of fused image for discriminating the object classes at 23.5 m spatial and 10-bit radiometric resolutions. The merged data set gives the improved classification accuracy.     

The Effect of Radiometric Resolution on the Retrieval of Leaf Area Index from Agricultural Crops

Motivated by the increasing availability of remote sensing data of high radiometric resolution, a study was conducted to determine whether high-resolution data (10 bits or more) yielded more accurate vegetation leaf area index (LAI) information than low-radiometric-resolution multispectral data (7-bit or less). The study evaluated the performance of simulated 12-bit LISS-III sensor data (derived from EO-1 hyperspectral Hyperion data) with original 7-bit LISS-III sensor data for the estimation of LAI of major agricultural crops (e.g., cotton, sugar cane, and rice). There was no significant improvement in the correlation coefficient encountered when using the high-radiometric-resolution (12-bit) LISS-III data versus the low-radiometric-resolution (7-bit) LISS-III data for the retrieval of LAI. The retrieval of LAI of agricultural crops met with moderate success, with overall correlation coefficients of around 0.55. These results suggest that satellite data of very high radiometric resolution may not be a required for remote measurement of LAI. Spectral bandwidth, band placement, and the method of retrieving biophysical parameters may be more important.     

高分辨率遥感影像林地边界提取方法

由于高分辨率遥感影像上的信息高度细节化,加之噪声的影响,会导致基于像元级纹理特征的林地边界提取方法的效果不理想。为此,提出一种基于种子纹理基元合并的半自动林地边界提取方法。首先利用基于图模型的影像分割算法获取初始基元;然后定义了一种针对非规则基元统计基元级灰度共生矩阵(GLCM)纹理特征的方法;最后在人工给定种子基元的基础上合并具有相似纹理的基元,并对基元合并的结果进行边界提取,得到高分影像上的林地边界。利用多源高分影像对所提方法进行验证及对比分析。实验结果表明,该方法对高分影像上大片典型林地的边界可取得较高的提取精度和计算效率。     

Comparison of sampling methods for inventorying the stand volume using satellite remote sensing

Remote sensing is being increasingly used for forest resource inventory as it saves time and the cost. Aerial photographs and satellite images have been effectively utilized for forest inventory all over the world. This study highlights the application of IRS LISS-III imagery for inventorying the stand volume in Lachchhiwala Forest Range of Siwaliks. The satellite image was visually interpreted for forest type and density stratification. Both random as well as stratified random sampling techniques were used to see their impact on the volume estimates. Field sampling was done in the plots of 0.1 ha size. The total growing stock in all types of forests in the study area was estimated to be 1.87 mill.m³, of which Sal Forest accounted for 1.32 mill.m³, Sal Mixed Forest for 0.09 mill.m³, Mixed Sal Forest for 0.08 mill.m³, Miscellaneous Forest for 0.06 mill.m³ and Forest Plantations for 0.02 mill.m³. The results were compared with an independent field-based inventory carried out by forest department. The two sampling methods were compared by ratioing of the mean of variance (gain in precision) and it was found that the timber volume estimates using stratified random sampling technique were 15 per cent more accurate than simple random sampling. The satellite image-based inventory using stratified random sampling was found to have about 90 per cent correspondence with the inventory done by the Forest Department.     

ASTER和TM/ETM+遥感数据融合监测土地覆盖变化

在人们纷纷选择IKONOS、QUICKBIRD、SPOT-5等高分辨率影像监测土地利用/覆盖变化之际,以北京海淀区为例,尝试采用Brovey变换和主成分分析(PCA)法融合ASTER、TM/ETM+中等分辨率影像,充分利用ASTER、TM/ETM+数据的多光谱和较高空间分辨率特性,挖掘其在土地覆盖变化监测中的潜力,为大规模监测土地利用/覆盖变化提供科学参考。研究将2003年ASTER多光谱3N、2、1波段与1999年ETM+PAN波段进行Brovey变换;1992年TM543与1999年ETM+PAN波段进行PCA融合,快速发现土地覆盖变化信息。经验证,变化发现精度达92.50%,符合项目精度要求。试验表明:在缺乏高分辨率影像的地区,选择价格相对便宜的AS-TER和TM/ETM+数据,采用Brovey变换和主成分分析(PCA)法进行融合,可有效监测土地覆盖变化,节约动态监测成本,二者具有很大的应用价值,值得推广。     

Modelling for forest growing stock assessment using satellite data — A case study

The satellite digital vegetation index data has been correlated with the forest growing stock by fitting linear regression models. The goodness of fit was tested. The analysis showed that the vegetation index which is the ratio of reflectance of vegetation in near infrared band to red wave band of electromagnetic spectrum is highly correlated to forest growing stock and the same can be used to predict the volume in remote forest areas for quick assessment purpose. Implications for future forest inventory are discussed.     

Above Ground Forest Phytomass Assessment in Southern Gujarat

Spectral modeling of above ground biomass (AGB) with field data collected in 48 field sites representing moist deciduous forest in Surat district is reported. Models were generated using LISS-III and MODIS data. The plot-wise field data was aggregated to MODIS pixel (250 m) using area weightages of forest/vegetation. The study reports that above ground phytomass varied from 6.13 t/ha to 389.166 t/ha while AGB phytomass estimated using area-weights for sites of 250×250 m, ranged from 5.534 t/ha to 134.082 t/ha. The contribution of bamboo in AGB has been found very high. The analysis indicated that the highest correlation between AGB phytomass and red band (R) of MODIS satellite data of October was (R²=0.7823) and R²=0.6998 with both NDVI of October data as well as NDVI_max. High correlation (R²=0.402) with IR band of February month was also found. The phytomass range obtained by using MODIS data varies from 0.147 t/ha to 182.16 t/ha. The mean biomass is 40.50 t/ha. Total biomass is 31.44 Mt. The mean Carbon density is 19.44 tC/ha in forest areas. The study is validation of region-wise spectral modeling approach that will be adopted for mapping vegetation carbon pool of the India under National Carbon Project of ISRO-Geosphere Biosphere Programme.     

Comparison of ETM+ and MODIS Data for Tropical Forest Degradation Monitoring in the Peninsular Malaysia

This study was undertaken the use of course and moderate spatial resolution remote sensing data to assess the forest degradation in the Peninsular Malaysia. Moderate Resolution Imaging Spectroradiometer (MODIS) imagery was used as coarse spatial resolution data, while Landsat Enhanced Thematic Mapper⁺ (ETM⁺) imagery was used as moderate spatial resolution to compare the accuracy. Geometric and radiometric correction and re-sampling were performed in pre-processing to enhance the analysis and results. Canopy fractional cover was used as an approach to assess the forest degradation in this study. Then, an optimum vegetation index was selected to apply on canopy fractional cover to enhance the detection of forest canopy damage. At the same time, accuracy assessment for the approach was referred to the location of Neobalanocarpus Heimii and correlate with global evapotranspiration rate. The forest degradation analysis was also applied and compared for all of the states in the Peninsular Malaysia. In conclusion, Landsat ETM⁺ imagery obtained higher accuracy compare to MODIS using canopy fractional cover approach for forest degradation assessment, and can be more broadly applicable to use for forest degradation investigation.     

利用遥感进行退耕还林成活率及长势监测方法的研究

本文以张家口退耕还林工程的新造经济林为例,提出了一种利用高分辨率遥感技术监测新造林成活率及长势的方法。主要采用面向对象的信息提取技术提取退耕地新造林的树冠信息。开发了基于树冠分布图的树冠因子提取程序,计算树冠因子,统计造林成活率,从而掌握新造林地的现状。最后,根据实际测量的数据进行误差检验,由遥感数据自动提取的树冠冠幅平均误差为:东西冠幅为0.337m,南北冠幅为0.433m;计算新造林成活率的精度达到了89.837%。为退耕还林工程科学,高效的管理及决策支持提供了依据。     

Comparison of MODIS vegetation continuous field — based forest density maps with IRS-LISS III derived maps

The study compared forest cover maps derived using coarse resolution vegetation continuous fields (MODIS VCF; 500m resolution) with the maps derived from medium resolution (24m; IRS LISS-III) data. The comparison of VCF, per cent tree cover product, for the years 2000 to 2004 with LISS III forest density class maps of 2001 and 2003 was carried out for two sites representing hilly (Uttarakhand) and undulating terrains (Madhya Pradesh). Slicing VCF to corresponding forest crown cover, i.e., 0–10%, 10–40%, 40–70% and >70% produced considerable difference in forest area estimates when compared to original LISS III derived crown cover area. The corresponding value range in VCF for 0–10% of actual forest cover were 0–31% and 0–25% in 2 sites respectively, and the respective limit was consistent at 1–20% when VCF range were sliced with respect to upscaled LISS III at 500m resolution. Similarly, all other class limits were also found through iterative process. These limits were similar, within a site, across five years. Spatial Kappa match between these two data indicated higher match in 40–70% class, and also in undulating site. When compared at same resolution, similar forest area cover estimated with weighted area upscaling gave closest match. The study is useful in knowing the usability and limits of VCF product, and utility of spatial Kappa.     

Geospatial modelling of biological richness in Sunderbans

The present study attempts to assess the biological richness in Sunderban Biosphere Reserve (SBR) using a three-pronged approach i.e. satellite image (IRS 1D LISS-III) for vegetation/land use stratification, landscape analysis for disturbance regimes assessment and the disturbance regimes together with the ecosystem uniqueness, species richness and importance value for biological richness modelling. The study showed that four mangrove categories, viz., Avicennia, Phoenix, mixed mangroves and mangrove scrub, cover 23.21 per cent of the total geographical area of SBR. The largest area is occupied by mixed mangroves (18.31%). The overall accuracy of the vegetation/land use map worked out to be 91.67 per cent. The disturbance analysis revealed that the vegetation types were not much disturbed. Shannon-Weaver’s index of diversity was highest in case of mixed mangrove. The results revealed that 75 per cent forest area has high biological richness.     

Forest cover monitoring using remote sensing and GIS — A case study in Dhaulkhand range of Rajaji National Park,Uttar pradesh

Rajaji National Park in U.P. is a protected area where large number of nomad population live within the park area. Their dependence on the forest for cattle rearing and firewood has caused degradation of the forests. Proximity to settlements outside the park further adds to the problems. In the present study, forest cover and river, bed changes have been attempted by using temporal aerospace data of the year 1960 and 1993. Subsequently, PAMAP GIS package has been used for the change detection analysis. The study indicates that the land cover changes are mainly due to biotic factors. Some of the important changes in forest cover are: (i) transformation of mixed forest to scrub forest in 67 ha (ii) the sal mixed and mixed sal forest categories have replaced sal forest in 262 ha (iii) significant increase in forest plantation i.e. 2075 ha in the year 1960 to 3793 ha in 1993 (iv) eighteen times increase in Chirpine area, it increased from 13 ha in 1960 to 230 ha in 1993. The land cover changes in 6663 ha (45%) out of 14962 ha of the study area. The consequential changes in the river beds due to the change in the forest cover was also analyzed. During this period river beds with boulders have increased by 87 ha. These river bed changes include lengthening and broadening of river and change in river course.     

Multiresolution and Multispectral Data Fusion Using Discrete Wavelet Transform with IRS Images: Cartosat-1, IRS LISS III and LISS IV

Image fusion techniques integrate complimentary information from multiple image sensor data such that the new images are more suitable for the purpose of human visual perception and computer based processing tasks for extraction of detail information. As an important part of image fusion algorithms, pixel-level image fusion can combine spectral information of coarse resolution imagery with finer spatial resolution imagery. Ideally, the method used to merge data sets with high-spatial and highspectral resolution should not distort the spectral characteristics of the high-spectral resolution data. This paper describes the Discrete Wavelet Transform (DWT) algorithm for the fusion of two images using different spectral transform methods and nearest neighbor resampling techniques. This research paper investigates the performance of fused image with high spatial resolution Cartosat-1(PAN) with LISS IV and Cartosat-1(PAN) sensor images with the LISS III sensor image of Indian Remote Sensing satellites. The visual and statistical analysis of fused images has shown that the DWT method outperforms in terms of Geometric, Radiometric, and Spectral fidelity.     

Wall-to-wall spatial prediction of growing stock volume based on Italian National Forest Inventory plots and remotely sensed data

Spatial predictions of forest variables are required for supporting modern national and sub-national forest planning strategies, especially in the framework of a climate change scenario. Nowadays methods for constructing wall-to-wall maps and calculating small-area estimates of forest parameters are becoming essential components of most advanced National Forest Inventory (NFI) programs. Such methods are based on the assumption of a relationship between the forest variables and predictor variables that are available for the entire forest area. Many commonly used predictors are based on data obtained from active or passive remote sensing technologies. Italy has almost 40% of its land area covered by forests. Because of the great diversity of Italian forests with respect to composition, structure and management and underlying climatic, morphological and soil conditions, a relevant question is whether methods successfully used in less complex temperate and boreal forests may be applied successfully at country level in Italy.For a study area of more than 48,657 km² in central Italy of which 43% is covered by forest, the study presents the results of a test regarding wall-to-wall, spatially explicit estimation of forest growing stock volume (GSV) based on field measurement of 1350 plots during the last Italian NFI. For the same area, we used potential predictor variables that are available across the whole of Italy: cloud-free mosaics of multispectral optical satellite imagery (Landsat 5 TM), microwave sensor data (JAXA PALSAR), a canopy height model (CHM) from satellite LiDAR, and auxiliary variables from climate, temperature and precipitation maps, soil maps, and a digital terrain model.Two non-parametric (random forests and k-NN) and two parametric (multiple linear regression and geographically weighted regression) prediction methods were tested to produce wall-to-wall map of growing stock volume at 23-m resolution. Pixel level predictions were used to produce small-area, province-level model-assisted estimates. The performances of all the methods were compared in terms of percent root mean-square error using a leave-one-out procedure and an independent dataset was used for validation. Results were comparable to those available for other ecological regions using similar predictors, but random forests produced the most accurate results with a pixel level R² = 0.69 and RMSE_% = 37.2% against the independent validation dataset. Model-assisted estimates were more precise than the original design-based estimates provided by the NFI.     

Forest site quality assessment by aerial photo-interpretation case study of sal forests

93 black & white panchromatic aerial photographs of 1972 on 1:20,000 scale were used for assessing the site quality of Sal forest of part of West Dehradun Forest Division, by both direct and indirect methods. The forests were stratified into 5 height and 5 density classes and according to species composition. For necessary data for the assessment of growing stock, field measurements were made on 54 sample plots of 0.1 hectare each. This provided direct indication of actual productive capacity of a particular site. According to indirect method, ground vegetation was sampled in 54 quadrats of 5m x 5m each and has been classified into statistically sound vegetation groupings by normal association analysis technique. These groups facilitated the standardisation of their photo perameters. The indicator value of the plants in a grouping was used to assess the suitability of the site. This was substantiated by super-imposing growing stock, present at that site. By combination of the direct and indirect methods, sites have been classified into 3 categories i.e., suitable, conditionally suitable and unsuitable. Suggestions have been made for management of conditionally suitable sites and the cause for unsuitability of sites are explored. Hence, interpretation of qualitative attributes of vegetation alongwith the measurements of quantitative parameters on aerial photographs have proved to be useful for site quality assessment.     

Segmentation and classification of high spatial resolution images based on Hölder exponents and variance

Pixel-based or texture-based classification technique individually does not yield an appropriate result in classifying the high spatial resolution remote sensing imagery since it comprises textured and non-textured regions. In this study, Hölder exponents (HE) and variance (VAR) are used together to transform the image for measuring texture. A threshold is derived to segment the transformed image into textured and non-textured regions. Subsequently, the original image is extracted into textured and non-textured regions using this segmented image mask. Afterward, extracted textured region is classified using ISODATA classification algorithm considering HE, VAR, and intensity values of individual pixel of textured region. And extracted non-textured region of the image is classified using ISODATA classification algorithm. In case of non-textured region, HE and VAR value of individual pixel is not considered for classification for significant textural variation is not found among different classes. Consequently, the classified outputs of non-textured and textured regions that are generated independently are merged together to get the final classified image. IKONOS 1 m PAN images are classified using the proposed algorithm, and the classification accuracy is more than 88%.     

Image matching as a data source for forest inventory – Comparison of Semi-Global Matching and Next-Generation Automatic Terrain Extraction algorithms in a typical managed boreal forest environment

Image matching is emerging as a compelling alternative to airborne laser scanning (ALS) as a data source for forest inventory and management. There is currently an open discussion in the forest inventory community about whether, and to what extent, the new method can be applied to practical inventory campaigns. This paper aims to contribute to this discussion by comparing two different image matching algorithms (Semi-Global Matching [SGM] and Next-Generation Automatic Terrain Extraction [NGATE]) and ALS in a typical managed boreal forest environment in southern Finland. Spectral features from unrectified aerial images were included in the modeling and the potential of image matching in areas without a high resolution digital terrain model (DTM) was also explored. Plot level predictions for total volume, stem number, basal area, height of basal area median tree and diameter of basal area median tree were modeled using an area-based approach. Plot level dominant tree species were predicted using a random forest algorithm, also using an area-based approach. The statistical difference between the error rates from different datasets was evaluated using a bootstrap method.Results showed that ALS outperformed image matching with every forest attribute, even when a high resolution DTM was used for height normalization and spectral information from images was included. Dominant tree species classification with image matching achieved accuracy levels similar to ALS regardless of the resolution of the DTM when spectral metrics were used. Neither of the image matching algorithms consistently outperformed the other, but there were noticeably different error rates depending on the parameter configuration, spectral band, resolution of DTM, or response variable. This study showed that image matching provides reasonable point cloud data for forest inventory purposes, especially when a high resolution DTM is available and information from the understory is redundant.     

European forest cover mapping with high resolution satellite data: The Carpathians case study

The aim of the study was to elaborate a methodology for forest mapping based on high resolution satellite data, relevant for reporting on forest cover and spatial pattern changes in Europe. The Carpathians were selected as a case study area and mapped using 24 Landsat scenes, processed independently with a supervised approach combining image segmentation, knowledge-based rules to extract a training set and the maximum likelihood decision rule. Validation was done with available very high resolution imagery. Overall accuracies per scene ranged from 93 to 96%. The labelling disagreement in overlapping areas of adjacent scenes was 6.8% on average.