Semantic image segmentation has recently witnessed considerable progress by training deep convolutional neural networks (CNNs). The core issue of this technique is the limited capacity of CNNs to depict visual objects. Existing approaches tend to utilize approximate inference in a discrete domain or additional aides and do not have a global optimum guarantee. We propose the use of the multi-label manifold ranking (MR) method in solving the linear objective energy function in a continuous domain to delineate visual objects and solve these problems. We present a novel embedded single stream optimization method based on the MR model to avoid approximations without sacrificing expressive power. In addition, we propose a novel network, which we refer to as dual multi-scale manifold ranking (DMSMR) network, that combines the dilated, multi-scale strategies with the single stream MR optimization method in the deep learning architecture to further improve the performance. Experiments on high resolution images, including close-range and remote sensing datasets, demonstrate that the proposed approach can achieve competitive accuracy without additional aides in an end-to-end manner.

Building change detection is useful for land management, disaster assessment, illegal building identification, urban growth monitoring, and geographic information database updating. This study proposes an automatic method that applies object-based analysis to multi-temporal point cloud data to detect building changes. The aim of this building change detection method is to identify areas that have changed and to obtain from-to information. In this method, the data are first preprocessed to generate two sets of digital surface models (DSMs), digital elevation models, and normalized DSMs from registered old and new point cloud data. Thereafter, on the basis of differential DSM, candidates for changed building objects are identified from the points in the smooth areas by using a connected component analysis technique. The random sample consensus fitting algorithm is then used to distinguish the true changed buildings from trees. The changed building objects are classified as “newly built”, “taller”, “demolished” or “lower” by using rule-based analysis. Finally, a test data set consisting of many buildings of different types in an 8.5 km² area is selected for the experiment. In the test data set, the method correctly detects 97.8% of buildings larger than 50 m². The accuracy of the method is 91.2%. Furthermore, to decrease the workload of subsequent manual checking of the result, the confidence index for each changed object is computed on the basis of object features.

Automatic extraction of ground points, called filtering, is an essential step in producing Digital Terrain Models from airborne LiDAR data. Scene complexity and computational performance are two major problems that should be addressed in filtering, especially when processing large point cloud data with diverse scenes. This paper proposes a fast and intelligent algorithm called Semi-Global Filtering (SGF). The SGF models the filtering as a labeling problem in which the labels correspond to possible height levels. A novel energy function balanced by adaptive ground saliency is employed to adapt to steep slopes, discontinuous terrains, and complex objects. Semi-global optimization is used to determine labels that minimize the energy. These labels form an optimal classification surface based on which the points are classified as either ground or non-ground. The experimental results show that the SGF algorithm is very efficient and able to produce high classification accuracy. Given that the major procedure of semi-global optimization using dynamic programming is conducted independently along eight directions, SGF can also be paralleled and sped up via Graphic Processing Unit computing, which runs at a speed of approximately 3 million points per second.

High-resolution remote-sensing-image change detection is widely used in urban dynamic monitoring, geographic information updating, natural disaster monitoring, illegal building investigation, and land resource surveys. Common change-detection algorithms are mainly implemented in a fully supervised manner that relies on a large number of high-quality samples. Compared with a building change-detection dataset, a building semantic-segmentation dataset is easier to accumulate and obtain. Making full use of this semantic information in the design of a building change-detection network can effectively reduce the sample size required to train a change-detection model. In view of this, a semantic feature-guided Siamese change-detection framework is devised in this letter. The framework effectively exploits the prior information of building semantic features and uses the popular transformer structure to improve the change analysis module. The results of extensive experiments on two public datasets show that the framework is more accurate than the other state-of-the-art change detection algorithms and can effectively reduce the dependence of data on change detection samples in the model training process.

Triangulation is an important task in the 3D reconstruction of computer vision. It seems simple to find the position of a point in 3D space when its 2D perspective projections in multi-view images are given and the corresponding camera projection matrices are known. However, in practice, multiple lines in 3D space do not intersect at one point because of noise. Then how to calculate the optimal 3D point of intersection becomes difficult. While there have been multiple methods trying to solve this problem, there is no systematic comparison between them. In this paper, we reviewed various currently existing variants of triangulation method and compared them through extensive experiments. The speed and accuracy of these methods have been compared using both synthetic and real datasets. We presented the results of experi-ments and summarized the advantages, limitations, and applicability of these methods so that to provide a guide for users when they need to choose an appropriate triangulation method for their given applications. Moreover, based on above analysis we proposed an improved method which shows better performance.