Classification of Sonar Targets in Air: A Neural Network Approach

Ultrasonic sonar sensors are commonly used for contactless distance measurements in application areas such as automotive and mobile robotics. They can also be exploited to identify and classify sound-reflecting objects (targets), which may then be used as landmarks for navigation. In the presented work, sonar targets of different geometric shapes and sizes are classified with custom-engineered features. Artificial neural networks (ANNs) with multiple hidden layers are applied as classifiers and different features are tested as well as compared. We concentrate on features that are related to target strength estimates derived from pulse-compressed echoes. In doing so, one is able to distinguish different target geometries with a high rate of success and to perform tests with ANNs regarding their capabilities for size discrimination of targets with the same geometric shape. A comparison of achievable classifier performance with wideband and narrowband chirp excitation signals was conducted as well. The research indicates that our engineered features and excitation signals are suitable for the target classification task.

Advanced Sonar with Velocity Compensation

Robotics research is dependent on intelligent, fast, accurate, reliable, and cheap sensors. Sonar sensing can fulfill these requirements. Moreover, sonar physics provides robotics researchers with a natural selection capability for landmark detection in navigation problems. This paper presents a new sonar system that produces accurate measurement and on-the-fly single cycle classification of planes, corners, and edges. The paper shows how double pulse coding of the transmitted pulse can be exploited to simultaneously reject interference and perform classification. On a moving platform, the velocity of the sonar sensor affects range and bearing measurements and is dependent on the target type. These effects are analyzed in the paper. Effects of rotation are also considered. The analytical results derived in the paper for velocity-generated deviations in sonar range and bearing estimates are compared with experimental data. The experiments show that the deviations are evident at speeds encountered in mobile robot research. Compensation of these effects can be achieved using the analytical expressions derived in the paper, the sonar information of range, bearing and target classification, and odometry measurements.

Cirrhosis classification based on texture classification of random features

Accurate staging of hepatic cirrhosis is important in investigating the cause and slowing down the effects of cirrhosis. Computer-aided diagnosis (CAD) can provide doctors with an alternative second opinion and assist them to make a specific treatment with accurate cirrhosis stage. MRI has many advantages, including high resolution for soft tissue, no radiation, and multiparameters imaging modalities. So in this paper, multisequences MRIs, including T1-weighted, T2-weighted, arterial, portal venous, and equilibrium phase, are applied. However, CAD does not meet the clinical needs of cirrhosis and few researchers are concerned with it at present. Cirrhosis is characterized by the presence of widespread fibrosis and regenerative nodules in the hepatic, leading to different texture patterns of different stages. So, extracting texture feature is the primary task. Compared with typical gray level cooccurrence matrix (GLCM) features, texture classification from random features provides an effective way, and we adopt it and propose CCTCRF for triple classification (normal, early, and middle and advanced stage). CCTCRF does not need strong assumptions except the sparse character of image, contains sufficient texture information, includes concise and effective process, and makes case decision with high accuracy. Experimental results also illustrate the satisfying performance and they are also compared with typical NN with GLCM.

Wall-corner classification using sonar: a new approach based on geometric features

Ultrasonic signals coming from rotary sonar sensors in a robot gives us several features about the environment. This enables us to locate and classify the objects in the scenario of the robot. Each object and reflector produces a series of peaks in the amplitude of the signal. The radial and angular position of the sonar sensor gives information about location and their amplitudes offer information about the nature of the surface. Early works showed that the amplitude can be modeled and used to classify objects with very good results at short distances—80% average success in classifying both walls and corners at distances less than 1.5 m. In this paper, a new set of geometric features derived from the amplitude analysis of the echo is presented. These features constitute a set of characteristics that can be used to improve the results of classification at distances from 1.5 m to 4 m. Also, a comparative study on classification algorithms widely used in pattern recognition techniques has been carried out for sensor distances ranging between 0.5 to 4 m, and with incidence angles ranging between 20° to 70°. Experimental results show an enhancement on the success in classification rates when these geometric features are considered.

Fractional Fourier transform pre-processing for neural networks and its application to object recognition

This study investigates fractional Fourier transform pre-processing of input signals to neural networks. The fractional Fourier transform is a generalization of the ordinary Fourier transform with an order parameter a. Judicious choice of this parameter can lead to overall improvement of the neural network performance. As an illustrative example, we consider recognition and position estimation of different types of objects based on their sonar returns. Raw amplitude and time-of-flight patterns acquired from a real sonar system are processed, demonstrating reduced error in both recognition and position estimation of objects.

Neural networks for improved target differentiation and localization with sonar

This study investigates the processing of sonar signals using neural networks for robust differentiation of commonly encountered features in indoor robot environments. Differentiation of such features is of interest for intelligent systems in a variety of applications. Different representations of amplitude and time-of-flight measurement patterns acquired from a real sonar system are processed. In most cases, best results are obtained with the low-frequency component of the discrete wavelet transform of these patterns. Modular and non-modular neural network structures trained with the back-propagation and generating-shrinking algorithms are used to incorporate learning in the identification of parameter relations for target primitives. Networks trained with the generating-shrinking algorithm demonstrate better generalization and interpolation capability and faster convergence rate. Neural networks can differentiate more targets employing only a single sensor node, with a higher correct differentiation percentage (99%) than achieved with previously reported methods (61-90%) employing multiple sensor nodes. A sensor node is a pair of transducers with fixed separation, that can rotate and scan the target to collect data. Had the number of sensing nodes been reduced in the other methods, their performance would have been even worse. The success of the neural network approach shows that the sonar signals do contain sufficient information to differentiate all target types, but the previously reported methods are unable to resolve this identifying information. This work can find application in areas where recognition of patterns hidden in sonar signals is required. Some examples are system control based on acoustic signal detection and identification, map building, navigation, obstacle avoidance, and target-tracking applications for mobile robots and other intelligent systems.