Separating halftones from text is an important step in document analysis. We present an algorithm that accurately extracts halftones from other information in printed documents. We treat halftone extraction as a textu...
详细信息
Separating halftones from text is an important step in document analysis. We present an algorithm that accurately extracts halftones from other information in printed documents. We treat halftone extraction as a texture-segmentation problem. We show that commonly used halftones, consisting of a pattern of dots, can be viewed as a texture. This texture exhibits a distinct spectral component that can be detected using a properly-tuned Gabor filter. The Gabor (1946) filter essentially transforms halftones into high-contrast regions that can be isolated by thresholding. We propose a filter-design procedure and provide experimental results.
A multi-stage decimation filter is described which consists of comb filters and a single-stage finite impulse response(FIR) filter.A multi-rate comb filter structure is used to reduce the data path width. A multi-stag...
详细信息
A multi-stage decimation filter is described which consists of comb filters and a single-stage finite impulse response(FIR) filter.A multi-rate comb filter structure is used to reduce the data path width. A multi-stage interpolation filter consists of a half-band filter and a short FIR *** short FIR filter is designed with a nonlinear phase to realize an overall constant group delay signal *** coefficients of the FIR filters are optimally quantized to canonical-signed -digit form to realize multiplier-free(shift-and -add) filter implementations.
This paper presents an object-oriented NBO(Node-Block-Object) data model for hypermedia *** takes advantage of object-oriented method,encapsulates all multimedia information as well as link functions in one *** has su...
详细信息
This paper presents an object-oriented NBO(Node-Block-Object) data model for hypermedia *** takes advantage of object-oriented method,encapsulates all multimedia information as well as link functions in one *** has successfully achieved cross link to offer much better flexibility and two-way link to realize forward and backward searching in hypermedia system navigation.A conditional relation on links has also been realized, that is very helpful for time sensitive multimedia information processing and multimedia object cooperation.
With the fast development of multimedia technology,efficient management of multimedia data has become more and more *** this paper,a multimedia database is established. It has used hypermedia and object-oriented *** t...
详细信息
With the fast development of multimedia technology,efficient management of multimedia data has become more and more *** this paper,a multimedia database is established. It has used hypermedia and object-oriented *** that,to realize intelligent retrieval such as semantic retrieval,I add a simple expert system into *** can manage multimedia information such as sounds,images, texts and videos efficiently.
Although research into blind equalisation algorithms has been an on-going activity for more than a decade,the existing algorithms are often inefficient in combating the impairments introduced by rapidly, changing comm...
详细信息
Although research into blind equalisation algorithms has been an on-going activity for more than a decade,the existing algorithms are often inefficient in combating the impairments introduced by rapidly, changing communication *** and efficient algorithms are thus *** this paper,an efficient blind equalisation algorithm is proposed by modifying the existing Super Exponential *** method,based on fast Kalman filtering theory,is recursive in order and in time,and leads to a significant reduction in the number of arithmetic *** the order update by partitioning the covariance matrix of the first hundred data in a specific form,fast initialisation is also *** proposed algorithm requires computational complexity proportional to N but acquires the fast convergence characteristics of the Super Exponential *** results indicate that this equaliser performs better with a soft decision device applied at the output of the equaliser for various modulation schemes and channel conditions.
In this paper a novel feature extraction technique based on the two-dimensional DCT(Discrete Cosine Transform) and zigzag scanning of the spectrogram is *** is in contrast to conventional approaches based on single di...
详细信息
In this paper a novel feature extraction technique based on the two-dimensional DCT(Discrete Cosine Transform) and zigzag scanning of the spectrogram is *** is in contrast to conventional approaches based on single dimension analysis such as LPC,Cepstral,or FFT. As a phoneme recognition task,a series of experiments were conducted on the voice stops ('b','d','g')of the TIMIT[1]database uttered by 630 speakers(male & female).The extracted data form the basis for input patterns for training two types of neural networks,the semi-dynamic network(TDNN),and a static network(MLP). The highest recognition rates of 77.5 and 72.4 percent were recorded for TDNN and MLP *** contrasts with results of 72 percent quoted by Hwang et al[2]for the same phonemes spoken by 40 females.
In this paper VLSI-array architecture for the Hierarchical block-matching algorithm using a Mean Pyramid structure is *** to the novel arrangement of data flow,we can map HBMA onto systolic array without hardware redu...
详细信息
ISBN:
(纸本)0780329120
In this paper VLSI-array architecture for the Hierarchical block-matching algorithm using a Mean Pyramid structure is *** to the novel arrangement of data flow,we can map HBMA onto systolic array without hardware redundancy. The architecture can estimate motion vectors of 720×480 image with frame rate 30,for the displacement of±28 pixels,at 50MHz clock rate. We simulated the proposed architecture using Verilog-XL and synthesized it using Compass. Simulation results show that the architecture can be fabricated with the state-of-the-art CMOS technologies in one chip.
Ultrasonic non-destructive inspection is today widely employed in industries for detecting embedded or surface defects in engineering structures. However,the processing of data from ultrasonic inspection is rarely ***...
详细信息
Ultrasonic non-destructive inspection is today widely employed in industries for detecting embedded or surface defects in engineering structures. However,the processing of data from ultrasonic inspection is rarely *** objective of our study is to develop a dataprocessing system which makes use of image processing techniques in order to assist ultrasonic inspection operator in his difficult task. The research work described in this paper has thus been done with this objective in mind and it concerns a method that we have developed for the automatic-detection and characterization of *** method is based on the Hough Transform and it allows to localize defects in the component under inspection by an automatic analysis of the "ultrasonic" *** defects that are considered are tiny cracks in thick metallic structures.
In recent years,neural networks have been developed and widely used in oil and gas *** the neural network with traditional seismic prospecting methods,we put forward a new method to predict oil and gas reservoir and c...
详细信息
In recent years,neural networks have been developed and widely used in oil and gas *** the neural network with traditional seismic prospecting methods,we put forward a new method to predict oil and gas reservoir and calculate the thickness of *** method is applied in oilfields and the results are satisfactory.
Multirate systems and filter banks have traditionally played an important role in subband coding and compression for contemporary communication applications,and many of the key design issue in such applications have b...
详细信息
Multirate systems and filter banks have traditionally played an important role in subband coding and compression for contemporary communication applications,and many of the key design issue in such applications have been extensively *** this paper, the generation and implementation of multirate filter banks for Orthogonal Multicarrier Modulation(OMCM) transmission over multipath fading channel is addressed. We demonstrate that adaptive equalizer with the cross-terms between the neighbouring subchannel are necessary for implementation with small output errors,which is in critically subsamplied multiband *** our simulation,we also address the trade-off between the minimisation of program memory and data memory requirements in the compilation of dataflow programs for multirate signalprocessing.A practical example illustrating its efficiency is given.
暂无评论