Performance of Combination of Texture and Object Based Techniques in Image Classification for Urban Land Cover

<div><p class="IJARCSAbstract"><em>Satellite imagery paves way to obtain tangible information through remote sensing techniques. It is necessary to classify the image in order to extract the features. There exist various classification techniques and algorithms to retrieve various features from imagery. As the technology development proceeds in a faster track it is necessary to compensate its advancements by developing new techniques for feature retrieval. As far as high resolution satellite imagery are concerned object based feature retrieval and texture based feature retrieval techniques are gaining its importance. The texture based feature retrieval has various techniques involved in it, among which Haralick’s texture parameters has much importance. Thereby object based technique also has its own way of algorithms and processes for feature retrieval. The eCognition software provides a platform for combining texture and object based technique. It is well known from various journals that object based technique is best for classifying high resolution imagery. Thus the image is primarily segmented into objects for classification. The Haralick’s texture parameters which serve well in classification of urban land cover is chosen by computing statistical analysis. Finally the chosen texture parameter is adopted in the classification of the objects. The classified imagery is checked for accuracy and a high accuracy of 94.5% is obtained.</em></p></div>

Download Full-text

Object-based urban land cover classification using rule inheritance over very high-resolution multisensor and multitemporal data

GIScience & Remote Sensing ◽

10.1080/15481603.2015.1122923 ◽

2015 ◽

Vol 53 (2) ◽

pp. 164-182 ◽

Cited By ~ 18

Author(s):

Ejaz Hussain ◽

Jie Shan

Keyword(s):

High Resolution ◽

Land Cover ◽

Land Cover Classification ◽

Urban Land ◽

Urban Land Cover ◽

Object Based ◽

Multitemporal Data ◽

Very High

Download Full-text

WorldView-2 Data for Hierarchical Object-Based Urban Land Cover Classification in Kigali: Integrating Rule-Based Approach with Urban Density and Greenness Indices

Remote Sensing ◽

10.3390/rs11182128 ◽

2019 ◽

Vol 11 (18) ◽

pp. 2128 ◽

Cited By ~ 3

Author(s):

Mugiraneza ◽

Nascetti ◽

Ban

Keyword(s):

High Resolution ◽

Land Cover ◽

Informal Settlements ◽

Urban Land ◽

Urban Environments ◽

Urban Density ◽

Rule Based ◽

Urban Land Cover ◽

Object Based ◽

Rule Based Approach

The emergence of high-resolution satellite data, such as WorldView-2, has opened the opportunity for urban land cover mapping at fine resolution. However, it is not straightforward to map detailed urban land cover and to detect urban deprived areas, such as informal settlements, in complex urban environments based merely on high-resolution spectral features. Thus, approaches integrating hierarchical segmentation and rule-based classification strategies can play a crucial role in producing high quality urban land cover maps. This study aims to evaluate the potential of WorldView-2 high-resolution multispectral and panchromatic imagery for detailed urban land cover classification in Kigali, Rwanda, a complex urban area characterized by a subtropical highland climate. A multi-stage object-based classification was performed using support vector machines (SVM) and a rule-based approach to derive 12 land cover classes with the input of WorldView-2 spectral bands, spectral indices, gray level co-occurrence matrix (GLCM) texture measures and a digital terrain model (DTM). In the initial classification, confusion existed among the informal settlements, the high- and low-density built-up areas, as well as between the upland and lowland agriculture. To improve the classification accuracy, a framework based on a geometric ruleset and two newly defined indices (urban density and greenness density indices) were developed. The novel framework resulted in an overall classification accuracy at 85.36% with a kappa coefficient at 0.82. The confusion between high- and low-density built-up areas significantly decreased, while informal settlements were successfully extracted with the producer and user’s accuracies at 77% and 90% respectively. It was revealed that the integration of an object-based SVM classification of WorldView-2 feature sets and DTM with the geometric ruleset and urban density and greenness indices resulted in better class separability, thus higher classification accuracies in complex urban environments.

Download Full-text