Browsing by Author "Too, Boaz K."

Now showing 1 - 6 of 6
  • Thumbnail Image
    Item
    Extraction of Scene Text Information from Video
    (2016-01) Too, Boaz K.; Prabhakar, C. J.
    In this paper, we present an approach for scene text extraction from natural scene video frames. We assumed that the planar surface contains text information in the natural scene, based on this assumption, we detect planar surface within the disparity map obtained from a pair of video frames using stereo vision technique. It is followed by extraction of planar surface using Markov Random Field (MRF) with Graph cuts algorithm where planar surface is segmented from other regions. The text information is extracted from reduced reference i.e. extracted planar surface through filtering using Fourier-Laplacian algorithm. The experiments are carried out using our dataset and the experimental results indicate outstanding improvement in areas with complex background where conventional methods fail.
  • Thumbnail Image
    Item
    Extraction of Scene Text Information from Video
    (Modern Education and Computer Science, 2016) Too, Boaz K.; Prabhakar, C. J.
    In this paper, we present an approach for scene text extraction from natural scene video frames. We assumed that the planar surface contains text information in the natural scene, based on this assumption, we detect planar surface within the disparity map obtained from a pair of video frames using stereo vision technique. It is followed by extraction of planar surface using Markov Random Field (MRF) with Graph cuts algorithm where planar surface is segmented from other regions. The text information is extracted from reduced reference i.e. extracted planar surface through filtering using Fourier-Laplacian algorithm. The experiments are carried out using our dataset and the experimental results indicate outstanding improvement in areas with complex background where conventional methods fail.
  • Thumbnail Image
    Item
    Localization of Overlaid Text Based on Noise Inconsistencies
    (2015-04) Too, Boaz K.; Prabhakar, C. J.
    In this paper, we present a novel technique for localization of caption text in video frames based on noise inconsistencies. Text is artificially added to the video after it has been captured and as such does not form part of the original video graphics. Typically, the amount of noise level is uniform across the entire captured video frame, thus, artificially embedding or overlaying text on the video introduces yet another segment of noise level. Therefore detection of various noise levels in the video frame may signify availability of overlaid text. Hence we exploited this property by detecting regions with various noise levels to localize overlaid text in video frames. Experimental results obtained shows a great improvement in line with overlaid text localization, where we have performed metric measure based on Recall, Precision and f-measure.
  • Thumbnail Image
    Item
    Localization of Overlaid Text Based on Noise Inconsistencies
    (2015-04) Too, Boaz K.; Prabhakar, C. J.
    In this paper, we present a novel technique for localization of caption text in video frames based on noise inconsistencies. Text is artificially added to the video after it has been captured and as such does not form part of the original video graphics. Typically, the amount of noise level is uniform across the entire captured video frame, thus, artificially embedding or overlaying text on the video introduces yet another segment of noise level. Therefore detection of various noise levels in the video frame may signify availability of overlaid text. Hence we exploited this property by detecting regions with various noise levels to localize overlaid text in video frames. Experimental results obtained shows a great improvement in line with overlaid text localization, where we have performed metric measure based on Recall, Precision and f-measure.
  • Thumbnail Image
    Item
    Segmentation and Alignment of Multi-oriented and Curved Text Lines from Document Images
    (2015-06) Too, Boaz K.; Prabhakar, C. J.
    In this paper, we present a novel approach to segment and align multi-oriented and curved text-lines from document images. We assumed that the input document image contains text-lines with arbitrary orientation and identified the arbitrary text string based on projection profile. We employed anisotropic Gaussian filter bank on the identified arbitrary text region in order to smooth the text region, which helps to detect the ridges which is a representative of a text-line path. The ridges are then labeled and a cubic B-spline is fitted to the text-line path points. The orientation and curvature features of the text-line path is estimated using orientated gradients for each point and corresponding curvature to these text-line path are computed. Text is aligned along the horizontally transformed line by rotating individual characters based on the computed curvature information. Finally, the aligned text-lines are extracted, which can be fed into OCR for recognition. The evaluation metrics was evaluated at text-line segmentation level and the results posted show a significant improvement. The resulting system is proven to provide better results than most state of the art algorithms
  • No Thumbnail Available
    Item
    Staff Profile - Dr. Too Boaz Kipyego
    (2016-09) Too, Boaz K.
    Lecturer School of Pure and Applied Science, Mathematics, Computing & I.T. Department University of Embu