According to the UN Population Reference Bureau, 1.4 billion more people will have settled in urban areas by 2030. One of the key environmental effects of rapid urbanization is the urban heat island (UHI) effect. Understanding the mechanism of surface UHIs associated with land-use/land-cover (LULC) change patterns is important for improving the ecology and sustainability of cities. In this article, time series Landsat Thematic Mapper (TM)/Enhanced Thematic Mapper Plus (ETM+) data were used to extract LULC data and land surface temperature (LST) data for the city of Jinan, China, from 1987 to 2011, a period during which the city experienced rapid urbanization. With the aid of a geographical information system (GIS) and remote sensing (RS) approach, the changes in this urban area’s LULC were explored, and the impact of these changes on the spatiotemporal patterns and underlying driving forces of the surface UHI effect were further quantitatively characterized. The results show that significant changes in land use and land cover occurred over the study period, with loss of farmland, forest, and shrub vegetation to urban use, leading to spatial growth of impervious surfaces. Consequently, the land surface characteristics and spatiotemporal patterns of the UHI have changed drastically. According to the seasonal and inter-annual variations in intensity of UHIs, mean differences in UHI intensity between city centre, peri-urban, and nearby rural areas were stronger during summer and spring and weaker during winter and autumn. Spatially, there were significant LST gradients from the city centre to surrounding rural areas. The city centre exhibited higher LSTs and remarkable variation in LSTs, while the surrounding rural areas exhibited lower LSTs and lower variation in LSTs. Moreover, the analysis of LSTs and indices showed that great differences of temperature even existed in a LULC type except for variations between different LULC types. In addition, a local-level analysis revealed that the intensity of the UHI effect is proportional to the size of the urban area, the population density, and the frequent occurrence of certain activities.

Monitoring heavy metal stress in crops via hyperspectral remote sensing technology is an effective way. A new approach, namely the ratio difference of autocorrelation function first derivative (RDA), is proposed to extract characteristic regions of maize leaves spectra for the initially identification on contaminated category of copper (Cu) and lead (Pb). Simultaneously, empirical mode decomposition (EMD) and power spectral density (PSD) are integrated to construct EMD-PSD to visually discrimination on Cu and Pb stress from frequency domain perspective. In our work, pot experiment contaminated by Cu and Pb were designed and carried out, and corresponding chemical data, chlorophyll and spectra of maize leaves were collected. Based on acquired spectra, RDA is used to obtain indicators and characteristic intervals of spectra, and then EMD-PSD is applied to obtain intrinsic mode functions (IMFs) from spectra and PSD maps. Through experimental analysis, the following conclusions are drawn: 1) the red edge and red shoulder region of spectra can be used as candidate on indicator to find the characteristic regions of spectra, and integrated intersection spectral intervals are considerable to distinguish Cu and Pb; 2) PSD maps extracted by EMD-PSD significantly can discriminate stress of Cu and Pb with three-dimensional visualization. This study takes the combination of spectral domain and frequency domain as the exploration point, the stress of Cu and Pb was distinguished by mutual verification based on spectra (group I and group II and 2014 experiment). In summary, the proposed method can identify the weak differences of spectra and distinguish the stress of Cu and Pb qualitatively, which provides a new perspective for the identification of heavy metal stress categories.

Strong geometric and radiometric distortions often exist in optical wide-baseline stereo images, and some local regions can include surface discontinuities and occlusions. Digital photogrammetry and computer vision researchers have focused on automatic matching for such images. Deep convolutional neural networks, which can express high-level features and their correlation, have received increasing attention for the task of wide-baseline image matching, and learning-based methods have the potential to surpass methods based on handcrafted features. Therefore, we focus on the dynamic study of wide-baseline image matching and review the main approaches of learning-based feature detection, description, and end-to-end image matching. Moreover, we summarize the current representative research using stepwise inspection and dissection. We present the results of comprehensive experiments on actual wide-baseline stereo images, which we use to contrast and discuss the advantages and disadvantages of several state-of-the-art deep-learning algorithms. Finally, we conclude with a description of the state-of-the-art methods and forecast developing trends with unresolved challenges, providing a guide for future work.