Classification of plant species from CTFM ultrasonic range data using a neural network

Harper, N.; McKerrow, P.J.

doi:10.1109/icnn.1995.487728

Cited by 11 publications

(5 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Several studies have examined potted foliage of different plant species with continuous-wave frequency-modulated (CWFM) [47][48][49][50] or wideband sonar [7,51]. All these studies found class-specific differences in echo properties which demonstrated that echo classification was possible and formulated hypotheses as to how these properties relate to foliage structure.…”

Section: Classification Of Large Random Targetsmentioning

confidence: 99%

Biosonar-inspired technology: goals, challenges and insights

Müller

Kuc

2007

Bioinspir. Biomim.

View full text Add to dashboard Cite

Bioinspired engineering based on biosonar systems in nature is reviewed and discussed in terms of the merits of different approaches and their results: biosonar systems are attractive technological paragons because of their capabilities, built-in task-specific knowledge, intelligent system integration and diversity. Insights from the diverse set of sensing tasks solved by bats are relevant to a wide range of application areas such as sonar, biomedical ultrasound, non-destructive testing, sensors for autonomous systems and wireless communication. Challenges in the design of bioinspired sonar systems are posed by transducer performance, actuation for sensor mobility, design, actuation and integration of beamforming baffle shapes, echo encoding for signal processing, estimation algorithms and their implementations, as well as system integration and feedback control. The discussed examples of experimental systems have capabilities that include localization and tracking using binaural and multiple-band hearing as well as self-generated dynamic cues, classification of small deterministic and large random targets, beamforming with bioinspired baffle shapes, neuromorphic spike processing, artifact rejection in sonar maps and passing range estimation. In future research, bioinspired engineering could capitalize on some of its strengths to serve as a model system for basic automation methodologies for the bioinspired engineering process.

show abstract

Section: Classification Of Large Random Targetsmentioning

confidence: 99%

Biosonar-inspired technology: goals, challenges and insights

Müller

Kuc

2007

Bioinspir. Biomim.

View full text Add to dashboard Cite

show abstract

“…These particular sonar sounds are well suited to convey detailed shape information that can be used for target recognition. Most work using CTFM sonar has concentrated on a biological approach, using neural networks for identification (Harper and McKerrow 1995;Gorman and Sejnowski 1988). However, the sensors essentially return the same information as the sampled time of flight sensors, although in our experience the data are less prone to noise.…”

Section: Introductionmentioning

confidence: 95%

Feature Extraction from a Broadband Sonar Sensor for Mapping Structured Environments Efficiently

Kao

Probert

2000

The International Journal of Robotics Research

View full text Add to dashboard Cite

In this paper, we describe the use of a broadband, frequencymodulated sonar sensor (a CTFM sonar) for generating maps of structured environments for robots and autonomous systems. The major advantage of using these sensors is that the information generated is reliable enough to extract the geometry and type of certain features with very few measurements; hence, maps can be generated rapidly. Although the technology of low-cost CTFM sonar has been available for some time, it has only very recently been considered seriously as a candidate sensor for navigation and mapping in spite of its superior performance. We describe the operation of the sensor and the use of both the range-orientation and amplitude-orientation profiles to extract information online and point targets in both smooth and rough environments. We demonstrate the method through generating a map for a room typical of a modern building (in fact, in a domestic house), using resolution of scanning increment from 1.8 deg down to 10.8 deg. We examine the robustness of the method as the number of measurements is reduced through introducing two quality metrics for each map.

show abstract

“…In that case, the previous equation takes the more general form: The modulation rate of the transmitted signal determines the frequency range of the beat signal, here chosen to be in the audible range. Following Dror, Zagaski, and Moss (1995) and Harper and McKerrow (1995), we have called the amplitude spectrum of the beat signal the CTFM sonar signature, since it corresponds to a one-dimensional range map in the direction of the sonar. In Kao and Probert (2000), it is referred to as the sonar image.…”

Section: Operation Of Frequency Modulation Sonarmentioning

confidence: 99%

“…Until recently, this technology has not been introduced into robotics and related technologies, but Kao and Probert (2000) showed that it is efficient for recognizing geometric features for localization. A number of workers have attempted to use similar sensors for classification using neural networks (Dror, Zagaski, and Moss 1995;Harper and McKerrow 1995;Dror et al 1996), using as input various formats of echo representation (time domain, frequency domain, and time frequency). In contrast, the classifier we describe in this paper uses statistical features of the echo signal, in particular the distribution of energy with time.…”

Section: Introductionmentioning

confidence: 99%

Classification of Textured Surfaces for Robot Navigation Using Continuous Transmission Frequency-Modulated Sonar Signatures

Politis¹,

Probert

2001

The International Journal of Robotics Research

View full text Add to dashboard Cite

Whereas in the past ultrasonic sensors have been largely used only to estimate the proximity of objects and the location and identification of primitive targets in a robot workspace, the development of biomimetic sonar has opened up new possibilities for their application. Broadband sonar echoes have sufficient resolution so that characteristics on reflection, especially geometry and texture, can be distinguished with only a few measurements. In this paper, we describe how a model of texture can be used to distinguish between a number of different surfaces using only a single measurement of each, showing results on a number of surfaces that might be considered typical pathways for a mobile robot, both those with periodicity in pattern and those with statistically homogeneous features. In particular, we consider textures corresponding to hard smooth floors, carpets and asphalts, and surfaces with a repeating pattern made up of tiles. Each random rough surface is modeled using an extension of the Kirchhoff approximation method describing the scattering of the acoustic wave on the surface while the periodic surfaces are modeled assuming distinctive reflections from the tile borders. The continuous transmission frequency-modulated sonar signature corresponding to each class is derived and compared with the experimental measurement. A set of features is extracted that exploits the differences between the surface models, and a hierarchical classification scheme is proposed for recognition.

show abstract

Classification of plant species from CTFM ultrasonic range data using a neural network

Cited by 11 publications

References 8 publications

Biosonar-inspired technology: goals, challenges and insights

Biosonar-inspired technology: goals, challenges and insights

Feature Extraction from a Broadband Sonar Sensor for Mapping Structured Environments Efficiently

Classification of Textured Surfaces for Robot Navigation Using Continuous Transmission Frequency-Modulated Sonar Signatures

Contact Info

Product

Resources

About