Ki Gon Nam

Optoelectronic difference-of-Gaussian wavelet transform system

A wavelet transform is efficient for multiresolution signal analysis. Difference-of-Gaussian (DOG) wavelets belong to a particular class of wavelets that extract the information of a specific frequency range in an image. Different DOG wavelets are produced by simple diffusion and subtraction processes by using diffusion network. We propose an optoelectronic DOG wavelet transform system using the point spread function of the combined system, which is similar to the Gaussian function. The experimental results show that the proposed system has potential applications to the systems that require DOG wavelet transforms in real time.

Diffusion neural network model for image-preprocessing

In this letter we propose a neural network model that performs the Gaussian operation efficiently by the diffusion process. Diffusion of an external spot excitation to the neighbouring pixels results in a Gaussian distribution. We apply this diffusion model to the DOG (difference of Gaussian) operation to detect the intensity changes in an image. In this model each cell has four fixed-weighted interconnections to the neighboring cells for a two-dimensional image. A different spatial frequency component can be obtained in each step of a sequential diffusion process. Therefore, the diffusion model is simpler and more efficient than the well-known LOG masking method. As far as we know, this is the only model for edge detection that can be implemented in hardware.<<ETX>>

Optoelectronic implementation of a diffusion neural network for edge detection

We investigate the feasibility of an optoelectronic edge detection system, using a diffusion neural network. The diffusion neural network performs the Gaussian operation efficiently by means of the diffusion process. We apply this in producing the difference-of-two-Gaussians function, which can detect the intensity changes of an image. This system is composed of a spatial light modulator, a two-dimensional image sensor array, and a computer. The processing of the system can be done at a rate of 30 frames/s, making it potentially applicable to systems that require edge detection of an image in real time.

Optoelectronic implementation of diffusion neural network for contour detection

In this paper we investigate a feasibility of an opto-electronic implementation of the diffusion neural network for contour detection. The diffusion neural network performs the Gaussian operation efficiently by the diffusion process. We apply this in producing the DOG (Difference of two Gaussian) functions, which can detect the intensity changes of the different spatial frequency components in an image. In the diffusion neural network each neuron has four connections with the four nearest neighbor neurons and a self-decay loop for a 2D image, and the connection weights are fixed-valued. Therefore the diffusion neural network is simpler and more efficient than LOG masking method in hardware or optical implementation. We implement the diffusion neural network opto-electronically using the point spread function of a spatial light modulator. This system is composed of a spatial light modulator, a 2D image sensor array, and a computer. The processing time of the system is very fast. Therefore the system has a potential applicability to the system that requires a real time processing of an image.