Neil W. Bergmann

An Arabic optical character recognition system using recognition-based segmentation

Abstract Optical character recognition (OCR) systems improve human–machine interaction and are widely used in many areas. The recognition of cursive scripts is a difficult task as their segmentation suffers from serious problems. This paper proposes an Arabic OCR system, which uses a recognition-based segmentation technique to overcome the classical segmentation problems. A newly developed Arabic word segmentation algorithm is also introduced to separate horizontally overlapping Arabic words/subwords. There is also a feedback loop to control the combination of character fragments for recognition. The system was implemented and the results show a 90% recognition accuracy with a 20 chars/s recognition rate.

An automatic image quality assessment technique incorporating higher level perceptual factors

We present an objective image quality assessment technique which is based on the properties of the human visual system (HVS). It consists of two major components: an early vision model (multi-channel and designed specifically for complex natural images), and a visual attention model which indicates regions of interest in a scene through the use of importance maps. Visible errors are then weighted, depending on the perceptual importance of the region in which they occur. We show that this technique produces a high correlation with subjective test data (0.93), compared to only 0.65 for PSNR. This technique is particularly useful for images coded with spatially varying quality.

Guest editorial CAPA'08 configurable computing: Configuring algorithms, processes, and architecture issue I: Configuring algorithms and processes

Embedded computing systems have become an essential part of modern life. Mobile phones are now handheld computers, the automotive industry needs hundreds of embedded processors to build a modern car, and around us are smart sensor systems for monitoring, security, and healthcare, to mention some of them. About 90% of computer applications are embedded systems, most of these are mobile, wireless consumer appliances that must be small in size, with very low power consumption and high performance. Advances in microelectronics have changed the design concepts of embedded application-specific processors. Reconfigurable computing using field programmable gate arrays (FPGAs) have been used for a longtime in embedded digital system design. The reconfiguring is migrating from the circuit level to the level of simple processors and algorithms. The focus in dedicated processor design is no longer merely digital design, it is more reconfigurable multiprocessing. Configurable parallel processing has many advantages. First, it replaces time-consuming digital design by programming of multiprocessors, making the design reprogrammable. Second, mapping algorithms directly onto configurable space of simple processors provides nearly the same efficiency as applicationspecific integrated circuits (ASICs). Third, the multiprocessor concept facilitates the building of energy-efficient systems using dynamic shutdown of unused processors. And last, the performance is scalable, and depends on the algorithmic design, on the number of processors involved, and not on the clock frequency of electrical circuits. Efficient mapping of algorithms onto hardware requires simultaneous study of algorithms and architecture: The algorithm must fit the architecture and the architecture must fit the algorithm. To achieve a best match, one has to be able to (re)configure the architecture as well the algorithms. Coupling this with the multiprocessing requirements where several computational processes are running on the same computational space of elementary processors and connections, and competing for computational resources (processors and connections), a system with dynamic reconfiguration controlled by an operating system is formed.

Single bit error correction implementation in CRC-16 on FPGA

Framing protocols employ cyclic redundancy check (CRC) to detect errors incurred during transmission. Generally whole frame is protected using CRC and upon detection of error, retransmission is requested. But certain protocols demand for single bit error correction capabilities for the header part of the frame, which often plays an important role in receiver synchronization. At a speed of 10 Gbps, header error correction implementation in hardware can be a bottleneck. This work presents a hardware efficient way of implementing CRC-16 over 16 bits of data, multiple bit error detection and single bit error correction on FPGA device.

Automatic Self-Reconfiguration of System-on-Chip Peripherals

A technique is presented which allows an FPGA-based reconfigurable system-on-chip to automatically and dynamically load hardware peripheral controllers and software device drivers depending on the systems automated identification of peripheral boards which are connected to the FPGA. The technique loads peripheral detection modules into peripheral controller slots at system startup, and after these modules identify the peripheral, the correct hardware controllers and software drivers are loaded.

QUKU: a two-level reconfigurable architecture

FPGAs have been used for prototyping of ASICs, for low-volume ASIC replacement and for systems requiring in-field hardware upgrades. However, the potential to use dynamic reconfiguration to adapt FPGA operation to changing application requirements has been hampered by slow reconfiguration times, and poor CAD tool support. In this paper, a new architecture, QUKU (pronounced cuckoo), is described which uses a coarse-grained reconfigurable PE array (CGRA) overlaid on an FPGA. The low-speed reconfigurability of the FPGA is used to optimize the CGRA for different applications, while the high-speed CGRA reconfiguration is used within an application for operator re-use. An FIR filter kernel has been implemented on QUKU and is shown to have performance which bridges the gap between softcore CPUs and custom FPGA filter circuits

System requirements for industrial wireless sensor networks

The employment of device monitoring, diagnosis and condition-based maintenance is one of the potential approaches for enhancing operational efficiency and reducing energy consumption of industrial machines. Industrial wireless sensor networks (IWSNs) for device monitoring are appealing to industry due to their inherent advantages compared with traditional wired systems, such as low cost, convenience of installation and re-location. In this paper, firstly the design requirements of IWSNs are outlined based on surveying application examples and commercial systems. Secondly the wireless protocol standards, current off-the-shelf wireless sensor platforms and prototypes developed by individual researchers for IWSNs are listed and compared. The paper then describes two techniques for improving the efficiency and effectiveness of industrial wireless sensors systems - on-sensor data processing, and a modified MAC protocol for improved real-time performance. Finally possible solutions to deal with particular requirement of IWSNs, including a system architecture and a novel protocol stack for a monitoring system are discussed.

Bundled data asynchronous multipliers with data dependent computation times

A novel asynchronous design method is introduced which combines the area efficiency of bundled data with data dependent computation time. The design of a 16/spl times/16 bit multiplier using this technique is explained and evaluated. Simulation results show that area time savings of 20% compared to an equivalent synchronous design can be achieved.

SEMAT — The Next Generation of Inexpensive Marine Environmental Monitoring and Measurement Systems

There is an increasing need for environmental measurement systems to further science and thereby lead to improved policies for sustainable management. Marine environments are particularly hostile and extremely difficult for deploying sensitive measurement systems. As a consequence the need for data is greatest in marine environments, particularly in the developing economies/regions. Expense is typically the most significant limiting factor in the number of measurement systems that can be deployed, although technical complexity and the consequent high level of technical skill required for deployment and servicing runs a close second. This paper describes the Smart Environmental Monitoring and Analysis Technologies (SEMAT) project and the present development of the SEMAT technology. SEMAT is a “smart” wireless sensor network that uses a commodity-based approach for selecting technologies most appropriate to the scientifically driven marine research and monitoring domain/field. This approach allows for significantly cheaper environmental observation systems that cover a larger geographical area and can therefore collect more representative data. We describe SEMATs goals, which include: (1) The ability to adapt and evolve; (2) Underwater wireless communications; (3) Short-range wireless power transmission; (4) Plug and play components; (5) Minimal deployment expertise; (6) Near real-time analysis tools; and (7) Intelligent sensors. This paper illustrates how the capacity of the system has been improved over three iterations towards realising these goals. The result is an inexpensive and flexible system that is ideal for short-term deployments in shallow coastal and other aquatic environments.

A recognition-based Arabic optical character recognition system

Optical character recognition systems improve human-machine interaction and are widely used in many government and commercial departments. After forty years of intensive research, OCR systems for most scripts are well developed. However, not for Arabic script. Since Arabic is a popular script, Arabic OCR systems should have great commercial value. Thus a recognition-based Arabic OCR system is proposed in this paper. It consists of the image acquisition, preprocessing, segmentation, character fragmentation, combination of character fragments, feature extraction, and classification. A signal is fed back to improve and determine the segmentation/recognition result. The system has been implemented and it has 90% recognition accuracy with a 20 chars/sec recognition rate.