Search results

Dimension reduction increases rates of skin detection a,but easily leads to duplication of color and non-color pixels ,meanwhile threshold setting brings a certain error[5].
Section IV experimental results and performance analysis data processing.
(a ) color distribution of CrCb plane (b) before median filtering (c) after median filtering Figure 2.2 The color distribution of CrCb plane and boundary extraction Direct least squares ellipse fitting skin model In computer vision research,there are two ways on Ellipse fitting based on known data[7], one is Hough transform .Time and space complexity of the algorithm increases greatly When the parameters is more than 2[8].Another method is least square fitting, its calculation is simple but the robustness is weak[9].

Assume the discrete data of signal ( )x t is 12, , , nx x x� with certain sampling frequency, the time-domain features are extracted as follows: Skewness: 3 1 1 N i i x N α = = ∑ (1) Waveform index: 1 / | | | | N rms f rms i i X S N X x X = = = × ∑ (2) Kurtosis index: 4 4 4 1 1 / N i rms i rms K x X X N β = = = ∑ (3) Autocorrelation coefficient: ( ) ( ) / (0) x x x n t R n t R ρ ∆ = ∆ (4) Where maxX is the max amplitude of signal, rmsX is the root mean square amplitude, N is the length of discrete data and 1 1 ( ) ( ) ( ) N n x n R n t x r x r n N n − = ∆ = + − ∑ , 0,1,2 , ( ) n M M N = � � (5) Then add peak values and the peak frequencies in the specific frequency intervals as the frequency-domain
And through the above description, the training process can be seen as a dimension reduction and clustering to high dimensional features and the diagnosis is a matching process using this lowdimensional information.

Other research uses traffic accident data for analysis [4-5].
The main major drawback for this analysis is that data is sparse and hard to collect.
Because in the peak hour, the flow of traffic and pedestrian are all near saturation, the speeds are already very low, and the degree of reduction space is not very large.

Orthogonal experimental design, a design method of multi-factor and multilevel, can not only reduce the number of tests regularly, but also combine complex experimental design and data analysis closely, then the main factors which affect the test results would be caught in a shorter time, and a scientific analysis of the main factors would be made so that making each test representative [3].
In the freeze-thaw cycle, the reducing of pore volume means that the total specimen expansion pressure reduction, in addition, the migration of moisture becomes relatively difficult as the structure becoming dense, and osmotic pressure increases correspondingly more lag, so the concrete specimens with carbonization can stand more times of freeze-thaw cycles.
Conclusions 1) It can not only reduce the amount of test but also intuitively analyze and discuss data by using orthogonal design method to study carbonation effect on compressive strength of cement-fly ash mortar subjected to freeze-thaw cycles. 2) Through orthogonal analysis, carbonization time is found to be the main factors affecting compressive strength of cement-fly ash mortar freeze-thaw cycles under freeze-thaw, following by water-cement ratio, the number of freeze-thaw cycles and mix amount of fly ash. 3) Carbonization improves the frost resistance of concrete in some extent, and, in a reasonable range, the longer carbonation time is, the stronger frost resistance capacity is.

Introduction Gadolinium in nanocrystalline state shows significant modifications in magnetic properties comparing with coarse-grained: for example, substantial reduction in the effective atomic magnetic moment and an increase of the coercivity [1], the Curie temperature shift to lower temperatures [2], strong induced magnetic anisotropy in the crystallites [3].
The resistance data were normalized to the zero field value [R(H=0)] for simplicity.
There is a discrepancy between the MR and magnetization data for all samples (Fig. 5 a,b).

The destruction of the node can be evidently seen from Figure 6, and drawn from the data of Table 1 U type flashboard node is larger bearing capacity than Groove type flashboard node in line with engineering practice .
Thus affecting the bearing capacity of the entire tower, causing a major security risk. 2 Through this data we can see that the arc length method is basically correct to calculate the model results, and which reflectes broadly in line with the actual. 3 The flashboard is connected with the branch pipe and board that is the weak parts of the component.
Width to thickness ratio of strut stable strength reduction [J].

As a potential structural material for high temperature applications, the tribological performance of Ti3AlC2 at high temperatures is very important data but is still less known.
All XPS data were collected on a PHI-5702 multifunctional photoelectron spectrometer, using Al-Ka X-ray as the excitation source at a pass energy of 29.4 eV and a resolution of ±0.2 eV, The peak of adventitious carbon (C 1s: 284.8 eV) was used for calibration.
An obvious reduction in wear rates was found from 400°C to 600°C.

Table 1 The result of the comparison experiment result sample extracting times extraction time (min) extracting rate (%) total flavonoid content g/100g A1 1 25 33.6 3.03 A2 1 25 13.3 2.98 B1 2 60×2 29.0 2.53 B2 2 60×2 11.0 2.30 C1 1 30 34.1 3.12 C2 1 30 13.9 3.01 D1 2 90×2 30.2 2.50 D2 2 90×2 11.2 2.31 According to the data in the table, the conclusion is clear that the ultrasonic extraction tech. is better than decoction method.
(2) Through a large number of experimental researches on various plants medicine extraction, statistical data show that ultrasonic extraction process is very significant and considerable than decoction extraction process.
(3) The advantage of energy conservation (over 60%) and emission reduction (low carbon) is gotten with the ultrasonic extraction technology.

This operations and maintenance costs, renewal costs, and the cumulative support can be resulted a consequential effect on total cost reduction of ownership.
This institute is primary federal entity for gathering, analyzing, and presenting data related to education building facilities in the United States and nations [2].
NCES method provides a framework for decision maker, schools facilities managers, and the public to identify a basic set of schools facilities data, including definitions that will meet their information requirements [2].

Algorithm Principle Benford Model A statistical law called Benford’s Law [5, 6], was introduced by Frank Benford in 1938 and predicts the frequency of appearance of the MSD for a broad range of natural and artificial data.
Algorithm flow is shown in Figure 2, Algorithm procedure is as follows: (1)We first make the Gaussian blur and dimensionality reduction processing for test images, then make 8×8 not-repeated block discrete cosine transformation of images’ RGB three channels for getting 8×8 block DCT coefficient matrix
Experimental Results and Discussion Experimental Results We test by using Matlab R2011b, the experimental data are from Columbia University Image Database [10].