China 
Africa 
Explore chapters and articles related to this topic
A study of poultry realtime monitoring and automation techniques
Published in Arun Kumar Sinha, John Pradeep Darsy, Computer-Aided Developments: Electronics and Communication, 2019
A. Arun Gnana Raj, S. Margaret Amala, J. Gnana Jayanthi
X. Dong and J. Xie [9] presents generic features to characterize a bird species based on bird vocalisation retrieval. The audio recordings of birds are first converted to spectrograms using short-time Fourier transform, then a ridge detection method is applied to the spectrogram for detecting points of interest. The designed retrieval system has five major procedures. 1) Spectrograms are created using STFT with a Hamming window of 512 samples (23 ms) and 50% window overlap. 2) In Spectral Ridge Detection stage the ridge pixels are detected by convolving each prepared spectrogram with four masks, one mask for each ridge direction. 3) In Feature Extraction stage the Temporal entropy, Frequency bin entropy and Histogram of counts of four ridge directions are all calculated. 4) Indexing is to pre-calculate features for improving performance of matching process on a large audio collection. Indexing has been done and the generated results are ranked based on similarity scores. 5) In retrieval stage matching is achieved by using K-NN (K = 1). K-NN only count the individual highest neighbour when finalizing the retrieved calls. The experimental results show that the proposed feature set can achieve 0.71 in term of retrieval success rate. The spectral ridge feature gives better results than the MFCCs feature sets.
Image Fusion
Published in Fathi E. Abd El-Samie, Mohiy M. Hadhoud, Said E. El-Khamy, Image Super-Resolution and Applications, 2012
Fathi E. Abd El-Samie, Mohiy M. Hadhoud, Said E. El-Khamy
The motivation for this transform arose from the need to find a sparse representation of functions that have discontinuities along lines [104–107]. The ridgelet transform belongs to the family of discrete transforms employing basis functions. To facilitate its mathematical representation, it can be viewed as a wavelet analysis in the Radon domain. The Radon transform is a tool for shape detection. The ridgelet transform is, primarily, a tool for ridge detection or shape detection of the objects in an image. The two-dimensional (2-D) continuous ridgelet transform in R2 can be defined through the introduction of the following basis function [104–107]: ψa,b,θ=a−1/2ψ(x1cosθ+x2sinθ−b)a
A low-cost video-based pavement distress screening system for low-volume roads
Published in Journal of Intelligent Transportation Systems, 2018
Figure 1 shows the overall workflow of the proposed VPADS system. Any camera mounted in the car front can be used to collect video data; however, such a setting can only capture the video data in an oblique view, where existing crack detection techniques may not be applicable. To improve the system stability, we propose a data processing workflow by first defining the RoI within the current lane that is subsequently used to determine whether crack or distress features exist or not. The process includes first running the Canny edge detector and probabilistic Hough transform to locate sharp edges/nodes. Unsupervised clustering can be implemented, so that the detected edges/nodes are grouped into individual cluster. Those cluster points located on the road markings (and close to the camera position) are then used to fit two individual lines based on random sample consensus (RANSAC). The RoI can then be defined by bounding the area of the two fitted lines. After that, a multi-scale ridge detection filter is convoluted within the RoI so as to detect any local minima (potential crack or distress features). Post data processing is required to remove those island local minima through using boundary contour analysis, so as to retain the distress features (i.e., crack, pothole, rutting, bleeding, etc.). Finally, the location of the video image scene with distress condition can be identified and stored inside a geographic information system (GIS), if geo-tagged video images are used.