Optical median filtering using threshold decomposition

Image distance transforms that use optical correlation

The image distance transform is a time-consuming global operation. We use optical correlation for the fast computation of binary image distance transforms. A joint transform correlator is proposed to implement the algorithm that accommodates the morphological threshold-decomposition concept. The proposed optical processor is programmable for computing Euclidean, city-block, and chessboard distance transforms in real time. Skeletonization by use of the distance transform and the rotation invariance of the Euclidean distance transform are demonstrated.

Illumination-invariant pattern recognition with joint-transform-correlator-based morphological correlation

The performance of nonlinear morphological correlation is investigated and compared with that of conventional linear correlation. In particular, the effects of illumination variations on the morphological correlation output are investigated in detail. The morphological correlation is shown to be invariant to uniform input-image illumination when the input-image illumination is higher than that of the reference. It also provides higher pattern discriminability, sharper peaks, and more-robust detection in the presence of salt-and-pepper noise than does the linear correlation. Computer-simulation results are provided.

Modified morphological correlation based on bit-map representations

Pattern recognition with high discrimination can be achieved with a morphological correlator. A modification of this correlator is carried out by use of a binary slicing process instead of linear thresholding. Although the obtained correlation result is not identical to the conventional morphological correlation, it requires fewer calculations and provides even higher discrimination. Two optical experimental implementations of this modified morphological correlator as well as some experimental results are shown.

Nonlinear morphological correlation: optoelectronic implementation

An optoelectronic implementation of the nonlinear morphological correlation by use of a threshold-decomposition technique and a joint transform correlator architecture is presented. This nonlinear morphological correlation provides improved image detection compared with standard linear optical pattern-recognition correlation methods. It also offers a more robust detection of low-intensity images in the presence of high-intensity patterns to be rejected.

Optical implementation of visible gray-image morphology with the visual-area-coding technique

We present a novel scheme of visible gray-image morphology with the visual-area-coding technique (VACT). The VACT is a technique of digitized analog-optical computing in which data are converted into visible coded patterns and processed with the visible form. Because the achievable operations in the VACT are identical to those of mathematical morphology, mathematical morphology is adapted to gray-image morphology with the VACT. Computer simulation and optical experiments of the several operations in mathematical morphology verify the correctness of the proposed technique. The processing capacity of the proposed method is estimated in terms of the space-bandwidth product.

Rank-order and morphological enhancement of image details with an optoelectronic processor

In all-optical processors, enhancement of image details is the result of high-pass filtering. We describe an optoelectronic processor in which detail enhancement results from the digitally calculated difference between an original input image and its low-pass filtered version. The low-pass filtering is realized through the rank-order median and the morphological opening and closing operations calculated by use of the optical convolver. It is shown that the normalized difference between the morphological white and black top hats enhances bright and dark image details analogously to the rank-order unsharp masking.

Optical threshold-decomposed Boolean-based gray-scale filter

Because of the advances made in optical technology, an extension can be made from the well-known stack filter to a new kind of gray-scale filter by using arbitrary Boolean functions instead of positive Boolean functions in the second step of a three-step process. Applications of these threshold-decomposed Boolean-based gray-scale filters in image processing include image smoothing and feature detection. A general optical implementation approach of nonlinear threshold-decomposed Boolean-based gray-scale filters is also described that utilizes area coding, gray scale correlation and lookup table thresholding techniques. Some experimental results are demonstrated.

Optically programmable two-channel sorting network based on a polarization- and area-encoded technique

We describe an optical cascadable polarization- and area-encoded sorting system that is capable of independently sorting two arrays of unipolar data, each in an arbitrary order specified by a corresponding instruction. The sorting instructions can be easily programmed by external optical signals. The system can be used for sorting a single array of bipolar data.

Optical asymmetrical median filtering using gray-scale convolution kernels

Morphological transformations such as erosion and dilation can be performed optically by convolving the binary input image with a binary kernel and then passing the result through an ideal thresholder-hardlimiter. Generalizations result from using gray-scale convolution kernels. It is shown that gray-scale kernels can be used to perform transformations similar to openings and closing with just one convolution instead of the ususal cascade of erosion and dilation normally required for these operations. Optical experimental results are presented.

Optical symbolic substitution for morphological transformations

An optical architecture that uses symbolic substitution to perform morphological transformations is proposed. It is shown how the four basic morphological transformation operations can be posed as symbolic substitution problems. Representative examples of the application of morphological transformations to image processing are given.

Nearest-neighbor median filter

A new generalized version of the median filter is proposed. The new version preserves clear definitions of image details while keeping the ability of removing the impulsive noise; it is still simple in principle and easy to use. The input-output relationship and noise-removing power of the new version are compared with those of the standard median filter.