Building Recognition Based on Sparse Representation of Spatial Texture and Color Features

Li, Bin; Sun, Fuqiang; Zhang, Yonghan

doi:10.1109/access.2019.2905304

Cited by 5 publications

(2 citation statements)

References 18 publications

(25 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It has the purpose of learning how to reproduce on output y its own input data x. Internally, the network has a hidden layer h describing a code used to represent the input. The network can be understood as containing two parts: the encoder having function h = f (x), and the decoder having function r = g (h); thus, the autoencoder can be described by the function g (f (x)) = r [55]. Figure 4 presents an example of autoencoder that reduces the dimensionality while maintaining the main information.…”

Section: ) Autoencodermentioning

confidence: 99%

A Deep Artificial Immune System to Detect Weld Defects in DWDI Radiographic Images of Petroleum Pipes

2019

View full text Add to dashboard Cite

In recent years, research focused on (semi)automatic radiographic inspection methods has gained more attention. The present work proposes a method for detecting defects in radiographic images of welded joints of oil pipes. Real condition images obtained by the double wall double image (DWDI) technique usually present a lower quality when compared with images traditionally considered in many studies reported in the literature. First, the proposed approach detects discontinuities in DWDI radiographic images of welded joints, and then, based on a hybrid paradigm encompassing artificial immune systems (AIS) and deep learning (DL), it classifies each discontinuity as 'defect' and 'non-defect'. The proposed method performs two phases in the AIS module: early classification (based on negative selection) and evolving classification (based on clonal selection). In both phases, the pattern recognition task is performed using a set of features extracted from each discontinuity through a detector genetically encoded into immune cells. As an attempt to improve the classification performance, DL models (AlexNet and autoencoders) are incorporated aiming to increase the number of extracted features. Experiments performed on a set of 727 discontinuities show that the proposed approach achieves an Fscore of 70.7%, outperforming each of its modules running by themselves: AlexNet with Fscore = 64.86% and AIS with Fscore = 66%. Considering the challenges imposed by real conditions on image acquisition-and the low rates of false negatives-, results demonstrate that the proposed approach can be used to assist in inspection works when dealing with DWDI images. INDEX TERMS Artificial immune systems, deep learning, radiographic images, discontinuities classification, defect detection.

show abstract

Section: ) Autoencodermentioning

confidence: 99%

A Deep Artificial Immune System to Detect Weld Defects in DWDI Radiographic Images of Petroleum Pipes

2019

View full text Add to dashboard Cite

show abstract

“…According to [73], segmenting an image into separate areas is a complex procedure. It is usually necessary to distinguish objects of complex shapes from the general background surrounding the building.…”

mentioning

confidence: 99%

Building Façade Style Classification from UAV Imagery Using a Pareto-Optimized Deep Learning Network

et al. 2022

View full text Add to dashboard Cite

The article focuses on utilizing unmanned aerial vehicles (UAV) to capture and classify building façades of various forms of cultural sites and structures. We propose a Pareto-optimized deep learning algorithm for building detection and classification in a congested urban environment. Outdoor image processing becomes difficult in typical European metropolitan situations due to dynamically changing weather conditions as well as various objects obscuring perspectives (wires, overhangs, posts, other building parts, etc.), therefore, we also investigated the influence of such ambient “noise”. The approach was tested on 8768 UAV photographs shot at different angles and aimed at very different 611 buildings in the city of Vilnius (Wilno). The total accuracy was 98.41% in clear view settings, 88.11% in rain, and 82.95% when the picture was partially blocked by other objects and in the shadows. The algorithm’s robustness was also tested on the Harward UAV dataset containing images of buildings taken from above (roofs) while our approach was trained using images taken at an angle (façade still visible). Our approach was still able to achieve acceptable 88.6% accuracy in building detection, yet the network showed lower accuracy when assigning the correct façade class as images lacked necessary façade information.

show abstract

Vision-Based System for Detection of Petrol Pump and Charging Station

Bhatlawande¹,

Joshi²,

Nagare³

et al. 2023

Lecture Notes in Networks and Systems

View full text Add to dashboard Cite

Building Recognition Based on Sparse Representation of Spatial Texture and Color Features

Cited by 5 publications

References 18 publications

A Deep Artificial Immune System to Detect Weld Defects in DWDI Radiographic Images of Petroleum Pipes

A Deep Artificial Immune System to Detect Weld Defects in DWDI Radiographic Images of Petroleum Pipes

Building Façade Style Classification from UAV Imagery Using a Pareto-Optimized Deep Learning Network

Vision-Based System for Detection of Petrol Pump and Charging Station

Contact Info

Product

Resources

About