Wiener biters for restoration of images and Wiener biters for imagerecognition are well established. We present the relations between two versions of Wiener restoration filters and develop new conceptual interpretati...
详细信息
Wiener biters for restoration of images and Wiener biters for imagerecognition are well established. We present the relations between two versions of Wiener restoration filters and develop new conceptual interpretations of the Wiener recognition filter, in the process of which interesting features are revealed. In one interpretation we show that the correlation signal resulting from Wiener recognition filtering may be regarded as the restoration of a hidden delta function. In another interpretation we regard the Wiener recognition filter as a two-filter scheme whereby an inverse filter and a Wiener restoration filter are operated sequentially. Results of simulations of the latter representation are presented for various input-noise models, providing useful insight into the implications of such an interpretation. (C) 1996 Optical Society of America
We describe a nonlinear joint transform correlator-based two-layer neural network that uses a supervised learning algorithm for real-time face recognition. The system is trained with a sequence of facial images and is...
详细信息
We describe a nonlinear joint transform correlator-based two-layer neural network that uses a supervised learning algorithm for real-time face recognition. The system is trained with a sequence of facial images and is able to classify an input face image in real time. Computer simulations and optical experimental results are presented. The processor can be manufactured into a compact low-cost optoelectronic system. The use of the nonlinear joint transform correlator provides good noise robustness and good image discrimination.
Analytic criterion for choice of the optimum phase in binary phase-only filters that will provide high correlation performance in optical imagerecognition is developed.
Analytic criterion for choice of the optimum phase in binary phase-only filters that will provide high correlation performance in optical imagerecognition is developed.
An optical-digital approach to fabricating matched spatial filters for multiple imagerecognition is discussed. Multiplexing is accomplished in the Fourier plane following optical transformation. A screening filter is...
详细信息
An optical-digital approach to fabricating matched spatial filters for multiple imagerecognition is discussed. Multiplexing is accomplished in the Fourier plane following optical transformation. A screening filter is formed digitally by adding or subtracting the intensity spectra of the transformed images to be recognized or ignored. This screening filter is then used to construct optically a composite image which contains the essential features of the various images to be recognized and from which compromising features have been deleted. The composite image is then used to record the desired filter. Initial results for both recognition and discrimination are presented.
This paper presents a comparison between the field of artificial neural network and the field of image processing and pattern recognition. It shows that some of the adaptive processing algorithms for pattern recogniti...
详细信息
This paper presents a comparison between the field of artificial neural network and the field of image processing and pattern recognition. It shows that some of the adaptive processing algorithms for pattern recognition and image processing, in terms of neural networks, can be seen as adaptive heteroassociative and autoassociative memories, respectively. The similarities and differences between these two fields are addressed.
A simulated annealing algorithm is introduced to encode binary phase-only filters optimally for imagerecognition. Similar patterns that cannot be distinguished with conventional filter encoding methods are clearly di...
详细信息
A simulated annealing algorithm is introduced to encode binary phase-only filters optimally for imagerecognition. Similar patterns that cannot be distinguished with conventional filter encoding methods are clearly distinguished with the optimized filter. The computational requirements for optimizing the filter are not excessive.
The feasibility of classification of stochastic images for color vision in real time has been investigated with two approaches. First, a hybrid incoherent optical correlator based on a quasi-monochromatic cathode ray ...
详细信息
The feasibility of classification of stochastic images for color vision in real time has been investigated with two approaches. First, a hybrid incoherent optical correlator based on a quasi-monochromatic cathode ray tube (CRT) is sequentially operated on red, green, and blue channels for statistical pattern recognition. Second, a color TV monitor is employed in the incoherent optical correlator to achieve spectral-spatial statistical pattern recognition in parallel in real time. The spectral-spatial statistical pattern recognition filter is designed with the least-squares linear mapping technique and is compensated for the smearing effects due to the finite spectral bandwidth of the light from CRTs. The experimental results provided demonstrate better recognition reliability when the spectral-spatial filters are used than when only spatial filters are employed.
暂无评论