With the development of intelligent transportation systems in recent years, automatic License Plate Detection and Recognition (LPDR) for automobiles has raised significant research interests. LPDR can be applied in many fields such as in traffic management and security surveillance. There has been a lot of work done on LPDR for regular scenarios. However, rare research can be found for complex traffic scenarios, such as night scenes, where the images or videos often appear to be blurred, too dark or bright, or the distance might be too far away or too close. These situations may pose great challenges to detection and recognition. In this paper, we established an automatic pipeline system for real-time LPDR in complex traffic scenes using DeblurGANv2 for image deblurring, YOLOv7 for license plate detection, LPRNet for license plate recognition. We used the CCPD dataset as well as a synthetic license plate dataset to train our models. Our results indicate that the system works well, and the deblurring module can significantly improve the results of LPDR.
The body length and weight are critical physiological parameters for fishes, especially eel-like fishes like swamp eel(Monopterusalbus).Fast and accurate measuring of body length is significant for swamp eel culturing as well as its resource investigation and protection. This paper presents an Android smart phone-based photogrammetry technology for measuring and estimating the length and weight of swamp eel. This method utilizes the feature that the ratio of lengths of two objects within an image is equal to that of in reality to measure the length of swamp eels. And then, it estimates the weight via a pre-built length-weight regression model. Analysis and experimental results have indicated that this method is a fast and accurate method for length and weight measurements of swamp eel. The cross-validation results shows that the RMSE (root-mean-square error) of total length measurement of swamp eel is0.4 cm, and the RMSE of weight estimation is 11 grams.
The increasing volume of industrial solid wastes presents a critical problem for the global environment. In the detection and monitoring of these industrial solid wastes, the traditional field methods are generally expensive and time consuming. With the advantages of quick observations taken at a large area, remote sensing provides an effective means for detecting and monitoring the industrial solid wastes in a large scale. In this paper, we employ an object-oriented method for detecting the industrial solid waste from HJ satellite imagery. We select phosphogypsum which is a typical industrial solid waste as our target. Our study area is located in Fuquan in Guizhou province of China. The object oriented method we adopted consists of the following steps: 1) Multiresolution segmentation method is adopted to segment the remote sensing images for obtaining the object-based images. 2) Build the feature knowledge set of the object types. 3) Detect the industrial solid wastes based on the object-oriented decision tree rule set. We analyze the heterogeneity in features of different objects. According to the feature heterogeneity, an object-oriented decision tree rule set is then built for aiding the identification of industrial solid waste. Then, based on this decision tree rule set, the industrial solid waste can be identified automatically from remote sensing images. Finally, the identified results are validated using ground survey data. Experiments and results indicate that the object-oriented method provides an effective method for detecting industrial solid wastes.
A new approach for determining the forest leaf area index (LAI) from a geometric-optical model inversion using multisensor observations is developed. For improving the LAI estimate for the forested area on rugged terrain, a priori information on tree height and the spectra of four scene components of a geometric-optical mutual shadowing (GOMS) model are extracted from airborne light-detection and ranging (LiDAR) data and optical remote sensing data with high spatial resolution, respectively. The slope and aspect of the study area are derived from digital elevation model data. These extracted parameters are applied in an inversion to improve the estimates of forest canopy structural parameters in a GOMS model. For the field investigation, a bidirectional reflectance factor data set of needle forest pixels is collected by combining moderate-resolution-imaging-spectroradiometer (MODIS) and multiangle-imaging-spectroradiometer (MISR) multiangular remote sensing observations. Then, forest canopy parameters are inverted based on the GOMS model. Finally, the LAI of the forest canopy of each pixel is estimated from the retrieved structural parameters and validated by field measurements. The results indicate that the accuracy of forest canopy LAI estimates can be improved by combining observations of passive multiangle and active remote sensors.
Leaf Area Index (LAI) is a key vegetation structural parameter in ecosystem. Our new approach is on forest LAI
retrieval by GOMS model (Geometrical-Optical model considering the effect of crown shape and Mutual Shadowing)
inversion using multi-sensor observations. The mountainous terrain forest area in Dayekou in Gansu province of China
is selected as our study area. The model inversion method by integrating MODIS, MISR and LIDAR data for forest
canopy LAI retrieval is proposed. In the MODIS sub-pixel scale, four scene components' spectrum (sunlit canopy, sunlit
background, shaded canopy and shaded background) of GOMS model are extracted from SPOT data. And tree heights
are extracted from airborne LIDAR data. The extracted four scene components and tree heights are taken as the a priori
knowledge applied in GOMS model inversion for improving forest canopy structural parameters estimation accuracy.
According to the field investigation, BRDF data set of needle forest pixels is collected by combining MODIS BRDF
product and MISR BRF product. Then forest canopy parameters are retrieved based on GOMS. Finally, LAI of forest
canopy is estimated by the retrieved structural parameters and it is compared with ground measurement. Results indicate
that it is possible to improve the forest canopy structural parameters estimation accuracy by combining observations of
passive and active remote sensors.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.