Non-edge Pixels Research Articles

Performance and performance variations of sub-1 THz detectors fabricated with 0.15 µm CMOS foundry process

Antenna-coupled field-effect transistors were integrated as multi-pixel (5×10) detector arrays for electromagnetic radiation between 550 and 600 GHz using commercial 0.15 µm CMOS process technology. Reported is a minimum optical noise-equivalent-power (NEP) of 43 pW/√Hz and a maximum (capacitive-loading-limited) optical responsivity of 970 V/W (both values averaged). An electrical NEP of 9 pW/√Hz is estimated. Inter-chip variations are analysed with a set of 15 samples showing a low standard deviation of less than 8% for both responsivity and NEP at the optimum operation point. Intra-chip variation is low for non-edge pixels. Both the very good NEP values and the low variations indicate that a cost-efficient CMOS process is well suitable for reliable fabrication of multi-pixel terahertz focal plane arrays.

A Method of Restoring the Shadow Distortion in a Bounded Book Scanned Image by Using Non-Edge Segment

When a thick bounded book is scanned on a flatbed scanner, the folding areas adjacent to the left page and right page would not be contacted on the glass. Therefore, the folding areas are darkened because the light reflected from the folding areas is weakened. In order to restore the shadow distortion in the folding areas, distortion profile should be firstly estimated. This paper proposes a method for estimating shadow distortion from the bounded book scanned image itself and for restoring the shadow distortion. The estimation of the shadow distortion can be done by analyzing the information of the non-edge segment which consists of non-edge pixels. The analysis of the change of the average luminance in the non-edge segments can estimate the characteristics of the shadow distortion.

Postprocessing in Block-Based Video Coding Based on a Quantization Noise Model

We present a model of quantization noise in block-coded videos with some assumptions in wavelet domain and propose a postprocessing method to reduce the quantization noise based on the model. A frame of video sequences is considered as a set of one-dimensional (1-D) horizontal and vertical signals. The quantization noise is considered as the sum of the blocking noise and the remainder noise. We model the blocking noise as an impulse or that along with a dispersed impulse at each block boundary in the wavelet domain. The validity of the blocking noise model is investigated. We also model the remainder noise as white Gaussian noise at non-edge pixels in the wavelet domain. Whether the model accommodates well to the remainder noise or not is also examined. The blocking noise is reduced by subtracting a profile, whose strength is adaptively estimated, at each block boundary from the coded signal. The remainder noise then is reduced by a soft-thresholding. We also propose a fast algorithm for the proposed method by approximating coefficients of shape profiles used in blocking noise reduction and inverse wavelet transform (WT) filters used in remainder noise reduction. The performance is evaluated for QCIF video sequences coded by H.263 TMN5 with quantization parameter (QP) in the range of 5--25 and is compared to that of the MPEG-4 verification model (VM) post-filter. Experimental results show that the proposed method yields not only PSNR improvement of maximum 0.5--dB over the VM post-filter but also subjective quality nearly free of the blocking artifact and edge blur.

Segmentation of petrographic images by integrating edge detection and region growing

A novel approach to segmenting petrographic images is proposed in this paper. A series of edge operators with various sizes of masks are first defined. By considering a larger neighborhood, the effects of noise or surface irregularities on edges are reduced. Color edges in an image are obtained by combining the edge operators and a color edge detection algorithm. A seeded region-growing algorithm is then used to segment the image based on the color edge information and the distances between edge-pixels and non-edge pixels. Seed regions are created automatically. These regions grow simultaneously. After all pixels in the image are labeled, the boundaries shared by two regions are checked. If a boundary is weak enough, it is eliminated and the corresponding two regions are merged. In the ultimate segmented map, each region whose size is large enough corresponds to a mineral grain in the image. This approach has been implemented in C++ under the Linux environment. Three sets of petrographic images were used to test the method.

Image vector quantization with variable dimension blocks and edge preserving cost function

This paper proposes a vector quantization scheme that uses variable vector dimensions and improves the perceived quality of the reconstructed images by using an edge based cost function. The cost function used for encoding is modified to take into account the number of edge pixels in a particular encoding block. For a better subjective quality, the product of the fraction of the edge pixels and the fraction of the non-edge pixels should be minimized. This is included in the additive cost function of VQ and minimized as an unconstrained problem. This edge based VQ technique reduces edge artifacts like jagging, ringing and also VQ distortions that happen near object contours, thus improving the overall subjective quality of the reconstructed image.

Automatic gradient threshold determination for edge detection

We describe a method to automatically find gradient thresholds to separate edge from nonedge pixels. A statistical model that is the weighted sum of two gamma densities corresponding to edge and nonedge pixels is used to identify a threshold. Results closely match human perceptual thresholds even under low signal-to-noise ratio (SNR) levels.

A pattern recognition approach to the detection of complex edges

We treat edge detection as a supervised pattern recognition problem, in which the edge is modeled as a linear combination of basis functions. We show both theoretically and analytically that our selection of Constant, Step and Gaussian as the basis functions is optimal in the sense of mininizing the RMS fitting error. The parameters of these basis functions are learned during the training phase, and the recognition phase uses these learned parameters to locate pixels that belong to edges. This modeling of edges is suitable to edges found in radiographs, which are more complex than step functions. Since edges may appear in any direction, we first show how to determine the gradient direction of the edge. We present the basis functions selected for modeling edges in bone radiographs and show both theoretically and experimentally that they are optimal in the sense that they are the most suitable combination for approximating the bone edge when compared to polynomials with more parameters. We then compute the relevant coefficients and parameters to be used for discriminating between edge and non-edge pixels.

Automatic multithreshold selection

A recursive technique for multiple threshold selection on digital images is described. Pixels are first classified as edge pixels or nonedge pixels. Edge pixels are then classified, on the basis of their neighborhoods, as being relatively dark or relatively light. A histogram of the graytone intensities is obtained for those pixels which are edge pixels and relatively dark and another histogram is obtained for those pixels which are edge pixels and relatively light. A threshold is selected corresponding to the graytone intensity value corresponding to one of the highest peaks from the two histograms. To get multiple thresholds, the procedure may be recursively applied first using only those pixels whose intensities are smaller than the threshold and then only those pixels whose intensities are larger than the threshold.