Kamran Reihani

Processing hierarchy for 2-D image structures

Abstract The processing hierarchy for 2-D image structures based on geometric considerations is presented. The proposed pattern-recognition system is based on a structural approach that couples symbolic and numerical methods to manipulate symbolic descriptions, namely spatial occurrences of lines, arcs, circular regions, etc., each with values, properties and constrained relationships to other symbolic entities. The processing hierarchy includes hierarchical modeling and symbolic manipulation of 2-D image structures—aspects of the image-formation process. In this paper, a hybrid strategy for designing a character-recognition system is proposed that contains embedded coupled numerical and symbolic processing units and identified raw data and stored models knowledge sources. Among the issues discussed is the process for evaluating image structures via computing dissimilarity levels (or variations) between feature descriptions of the prestored and extracted image structures. The variational concept has been implemented and the results are very encouraging.

Symbolic representation, modeling, and manipulation of arcs and lines

From a computational perspective, a general vision problem is considered by many to be an ill-posed problem. Currently, the dominant approach to characterizing vision computations is to categorize the required processings into low, intermediate, and high levels. This paper adopts a structural approach to vision that couples symbolic and numerical methods and discusses the semantics, modeling, and manipulation of symbolic descriptions of arc and line image structures for vision applications. Collectively, this approach demonstrates a logical perspective on the processes that occur in low, intermediate, and high levels of vision computations.© (1993) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Knowledge representation in a guided globular reasoning framework (Part II)

The processing necessary for high level vision is considered here. The manipulation of symbolic entities, each with values, properties, and constrained relationships to other symbolic entities is a major component. To perform recognition, the system operates in a guided globular reasoning framework. Further, concepts for a vision system which center on deduction are inherently distributed and desirable for an asynchronous systolic array architecture. The concepts presented in this paper, which are normally inaccessible, ease the tasks of processing spatial knowledge and modeling a modularly structured system.

Expert system modeling of a vision system

The proposed artificial intelligence-based vision model incorporates natural recognition processes depicted as a visual pyramid and hierarchical representation of objects in the database. The visual pyramid, with based and apex representing pixels and image, respectively, is used as an analogy for a vision system. This paper provides an overview of recognition activities and states in the framework of an inductive model. Also, it presents a natural vision system and a counterpart expert system model that incorporates the described operations.

Spiral systolic design with asynchronous controls for LPF digital filters

The design of spiral systolic arrays (Sas) with asynchronous controls for efficient and flexible implementation of linear phase FIR filters is presented. In this approach, (1) reduction of filter due to symmetry, (2) conversion of sequential input signal into input blocks by means of a spiral systolic mesh that is (a) suitable for highly parallel processing and (b) flexible for enabling various array sizes, and (3) making data streams independent of computations executed in each processor will collectively minimize computation time. The SA architecture processes the input signal row by row, and eliminates the complex shift register organization of the traditional FIR realization. Incorporated in this design are maximum parallelism and pipelinability, trade-off among computations, communications, and memory. Moreover, the systolic array will use simple local interconnection without undesirable properties such as preloading input data or global broadcasting. The key component of the asynchronous spiral SA is a communication protocol that controls input data flow properly and efficiently.

Systolic realization with asynchronous communication protocols for arbitrarily large LMS adaptive filters

The systolic algorithm approach is currently the most effective design procedure for minimizing computing time and the number of processors. In this study, a method is proposed for designing systolic arrays for arbitrarily large LMS filters. The development process involves the design of the (1) computational graph, (2) computational scheme, (3) interconnection scheme, (4) proposed processing element, for computing the output, yj at time j, and PEk protocol. The protocol controls input data flow properly and efficiently. This control ensures that values of input variables are not overwritten during their computing, while achieving maximum parallelism and thus reducing waiting time.

Flexible systolic design with asynchronous communication protocols for discrete Fourier transform (DFT) and inverse discrete Fourier transform (IDFT)

In this paper, a spiral systolic array (SA) architecture with asynchronous controls for the real time realization of an N point DFT and IDFT is considered. The study includes the overall system block diagram, propagation of data between array PEs, operations within each PE, and the PE protocol for computing an N point DFT. The idea is to design self-timed processors and communication protocols to gain control of data streams such that each computation can start if all its data are available, thus reducing waste time. The PE protocol controls input data flow properly and efficiently. The proposed SA has large throughput at the expense of more hardware than FFTs.

Learning in adaptive vision systems

The problem of learning in adaptive vision systems is considered here. This study explores the spatial and deductive reasoning aspects of the system by case-based techniques, as it computes (1) the invisibility justification of a birds eye, and (2) generalization and expansion of the gained knowledge in (1). The method presented manipulates symbolic entities, each with values, properties, and constrained relationships to other symbolic entities, to learn to create new knowledge from the existing knowledge.

Knowledge representation in a guided globular reasoning framework (part IV)

The study on the learning mechanisms of high-level vision systems can help us understand the link between spatial characteristics of objects and accumulated knowledge. This paper proposes a learning scheme that probes local objects that have experienced structural deviations, in conjunction with prestored knowledge, to compute conceptual elements or spatial knowledge. The newly developed conceptual elements can be used to model a modularly structured cognitive system.

Geometrical and morphological image processing algorithm

Early vision processing involves segmenting and eliminating/minimizing extrinsic variations in an image. Tradition segmentation algorithms use a single threshold to reduce image data to the regions important to the vision system. A single threshold rarely gives good segmentation results over an entire image because it may reject too many of the objects as belonging to the background or may accept too much of the background as belonging to the objects. This paper presents a segmentation method that uses multiple thresholds for partitioning an image into objects and background. The segmentation process is a gradient approach whereby the multiple (or adaptable) thresholds are calculated from discriminant structural feature values.