Skin is a widely used cue for porn image classification. Most conventional methods are off-line training schemes. They
usually use a fixed boundary to segment skin regions in the images and are effective only in restricted conditions: e.g.
good lightness and unique human race. This paper presents an adaptive online training scheme for skin detection which
can handle these tough cases. In our approach, skin detection is considered as a classification problem on Gaussian
mixture model. For each image, human face is detected and the face color is used to establish a primary estimation of
skin color distribution. Then an adaptive online training algorithm is used to find the real boundary between skin color
and background color in current image. Experimental results on 450 images showed that the proposed method is more
robust in general situations than the conventional ones.
On the basis of the mechanically scratching using the Atomic Force Microscope, this paper proposes a new method for manufacturing high frequency grating. The grating was fabricated on a polycarbonate compact disc with a silicon AFM tip under the contact mode. The fabrication technique and the optimization of parameters for the technique are discussed in detail. From the experiment, the minimum spacing of the grating can reach 30 nm. The digital nano-moire patterns verify that the grating has good potential to be applied to the nano-deformation measurement.
In this paper, an on-line down-well crude oil analysis system is presented, which is applied to measurement of the mount of components in down-well oil, water and gas. Instead of full spectrum of oil, the disperse-spectrum-selecting algorithm was presented to simplify the design of conventional spectrum analysis system. By the disperse-spectrum-selecting algorithm, the system can be designed in small size and compact structure, and the experiment results show that the correlation of the model using disperse wavelength is equivalent to the model using the all wavelengths and system measurement accurate reach to 2 percent.
A novel optical fiber null coupler (OFNC) sensor based on acousto-optic interaction is developed, which can be used in the structure health monitoring of the medical materials. The OFNC sensors can be response to 10MHz supersonic wave, and their signal-to noise ratio are higher then Piezo Ceramic Transducers(PZT). A kind of Perspex with a 1mm hole is employed as the sample, where the OFNC sensor is glued on, and the reflected signal of ultrasonic wave by the hole is detected .