Target Depth Information Research Articles

Target location method of intelligent deicing robot based on nonlinear auto disturbance rejection neural network

A visual navigation scheme for intermittent walking deicing robots was designed to address the issues of multiple iterations, long computation time, and poor real-time performance in the nonlinear optimization of the original SLAM system. A computer image acquisition unit, a computer image processing module, and a visual navigation parameter extraction algorithm were also designed. Implemented a visual navigation system for deicing robots based on image processing. It can meet the navigation requirements of the deicing robot. Simulation and experiments have shown that the method proposed in this paper can quickly and accurately identify abnormal point clouds. The visual servo control scheme constructed in this paper is aimed at robot task operations with unknown system calibration and target depth information. By calibrating the robot vision system, the conversion between camera pixel coordinates and robot base coordinates is achieved, with a transmission frame rate of 53.65 per second; The maximum error in positioning accuracy in space is 10.6 mm. The feature trajectory of visual space is smooth and stable within the camera's field of view, and the end movement of Cartesian space robots is stable without rebound, resulting in high grasping and positioning accuracy.

Moving Target Depth Information Extraction Based on Nonlinear Strategy Network

To improve the effect of depth information extraction of moving targets in the network, a nonlinear strategy-oriented method is proposed. With the advancement of science and technology, especially in wireless networks, a large amount of data is provided to people every hour of every day. Hence, it can increase the demand for data analysis tools. Nonlinear system modeling by using rough set theory to extract valuable information from large amounts of information, and then through the analytic hierarchy process (ahp) to determine the effect of input factors, then use particle swarm optimization algorithm (PSO) to find the accurate function, and USES the adaptive and population catastrophe and vaccine algorithm to make it to the local optimum, to achieve the aim of the complex. The experimental results show that, compared with M2 and M1 for 30 groups of samples, the model obtained by using M2 has a better fitting effect on the actual curve. The error of M2 is within ±3%, and the error of M1 is within ±6%, and the error is relatively large. The accuracy of the proposed method is higher than that of the neural network method, which proves that the nonlinear strategy is effective in the actual target depth information extraction.

FDA-SSD: Fast Depth-Assisted Single-Shot MultiBox Detector for 3D Tracking Based on Monocular Vision

In this study, a set of benchmarks for object tracking with motion parameters (OTMP) was first designed. The sample images were matched with the spatial depth of the camera, the pose of the camera, and other spatial parameters for the training of the detection model. Then, a Fast Depth-Assisted Single-Shot MultiBox Detector (FDA-SSD) algorithm suitable for 3D target tracking was proposed by combining the depth information of the sample into the original Single-Shot MultiBox Detector (SSD). Finally, an FDA-SSD-based monocular motion platform target detection and tracking algorithm framework were established. Specifically, the spatial geometric constraints of the target were adapted to solve the target depth information, which was fed back to the detection model. Then, the normalized depth information of the target was employed to select the feature window of the convolutional layer for the detector at a specific scale. This significantly reduces the computational power for simultaneously calculating detectors of all scales. This framework effectively combines the two-dimensional detection model and the three-dimensional positioning algorithm. Compared with the original SSD method, the network model designed in this study has fewer actual operating parameters; the measured detection operation speed was increased by about 18.1% on average; the recognition rate was maintained at a high level consistent with that of the original SSD. Furthermore, several groups of experiments were conducted on target detection and target space tracking based on monocular motion platforms indoors. The root mean square error (RMSE) of the spatial tracking trajectory was less than 4.72 cm. The experimental results verified that the algorithm framework in this study can effectively realize tasks such as visual detection, classification, and spatial tracking based on a monocular motion platform.

3D LIDAR imaging using Ge-on-Si single-photon avalanche diode detectors.

We present a scanning light detection and ranging (LIDAR) system incorporating an individual Ge-on-Si single-photon avalanche diode (SPAD) detector for depth and intensity imaging in the short-wavelength infrared region. The time-correlated single-photon counting technique was used to determine the return photon time-of-flight for target depth information. In laboratory demonstrations, depth and intensity reconstructions were made of targets at short range, using advanced image processing algorithms tailored for the analysis of single-photon time-of-flight data. These laboratory measurements were used to predict the performance of the single-photon LIDAR system at longer ranges, providing estimations that sub-milliwatt average power levels would be required for kilometer range depth measurements.

Time-delay Curve Detection and Target Depth Estimation in Ground Penetrating Radar Multi-layer Medium Detection

GPR is a useful tool in non-destructive evaluation area. Time delay curve detection and target depth estimation are two key steps in GPR data processing procedure, as they can provide accurate target depth information, especially in multi-layer medium detection scenario. Traditional methods for time-delay curve are applicable to homogeneous medium and ideal environments. This paper presents a new time-delay curve extracting approximation hyperbola algorithm (TDC-AH) and target depth algorithm. By applying auto-detecting edge points criterion in detection scenario, an approximation hyperbola equation can be established and target depth can be estimated subsequently. Simulating experiment proves the proposed algorithm is suitable for multi-layer medium.

Three-dimensional ghost imaging lidar via sparsity constraint.

Three-dimensional (3D) remote imaging attracts increasing attentions in capturing a target’s characteristics. Although great progress for 3D remote imaging has been made with methods such as scanning imaging lidar and pulsed floodlight-illumination imaging lidar, either the detection range or application mode are limited by present methods. Ghost imaging via sparsity constraint (GISC), enables the reconstruction of a two-dimensional N-pixel image from much fewer than N measurements. By GISC technique and the depth information of targets captured with time-resolved measurements, we report a 3D GISC lidar system and experimentally show that a 3D scene at about 1.0 km range can be stably reconstructed with global measurements even below the Nyquist limit. Compared with existing 3D optical imaging methods, 3D GISC has the capability of both high efficiency in information extraction and high sensitivity in detection. This approach can be generalized in nonvisible wavebands and applied to other 3D imaging areas.

Adaptive and robust algorithms and tests for visual-based navigation of a space robotic manipulator

Optical navigation for guidance and control of robotic systems is a well-established technique from both theoretic and practical points of view. According to the positioning of the camera, the problem can be approached in two ways: the first one, “hand-in-eye”, deals with a fixed camera, external to the robot, which allows to determine the position of the target object to be reached. The second one, “eye-in-hand”, consists in a camera accommodated on the end-effector of the manipulator. Here, the target object position is not determined in an absolute reference frame, but with respect to the image plane of the mobile camera. In this paper, the algorithms and the test campaign applied to the case of the planar multibody manipulator developed in the Guidance and Navigation Lab at the University of Rome La Sapienza are reported with respect to the eye-in-hand case. In fact, being the space environment the target application for this research activity, it is quite difficult to imagine a fixed, non-floating camera in the case of an orbital grasping maneuver. The classic approach of Image Base Visual Servoing considers the evaluation of the control actions directly on the basis of the error between the current image of a feature and the image of the same feature in a final desired configuration. Both simulation and experimental tests show that such a classic approach can fail when navigation errors and actuation delays are included. Moreover, changing light conditions or the presence of unexpected obstacles can lead to a camera failure in target acquisition. In order to overcome these two problems, a Modified Image Based Visual Servoing algorithm and an Extended Kalman Filtering for feature position estimation are developed and applied. In particular, the filtering shows a quite good performance if target's depth information is supplied. A simple procedure for estimating initial target depth is therefore developed and tested. As a result of the application of all the novel approaches proposed, the experimental test campaign shows a remarkable increase in the robustness of the guidance, navigation and control systems.

Depth-correction algorithm that improves optical quantification of large breast lesions imaged by diffuse optical tomography

Optical quantification of large lesions imaged with diffuse optical tomography in reflection geometry is depth dependence due to the exponential decay of photon density waves. We introduce a depth-correction method that incorporates the target depth information provided by coregistered ultrasound. It is based on balancing the weight matrix, using the maximum singular values of the target layers in depth without changing the forward model. The performance of the method is evaluated using phantom targets and 10 clinical cases of larger malignant and benign lesions. The results for the homogenous targets demonstrate that the location error of the reconstructed maximum absorption coefficient is reduced to the range of the reconstruction mesh size for phantom targets. Furthermore, the uniformity of absorption distribution inside the lesions improve about two times and the median of the absorption increases from 60 to 85% of its maximum compared to no depth correction. In addition, nonhomogenous phantoms are characterized more accurately. Clinical examples show a similar trend as the phantom results and demonstrate the utility of the correction method for improving lesion quantification.

The Mathematical Model and Applications of Coded Structured Light System for Object Detecting

According to the sensing principle of coded structured light, and the characteristics of broad complication and high parameters request, a new coded structured light mathematics model and measurement method is proposed, which can simply and quickly obtain depth information of targets. Model is simulated by computer. It can gain the offset and deformation of pattern images based on different model parameters. For the deformation of images, system parameters can be changed correspondingly. So, whole system can be fixed on using optimal system parameters. The experiment results reveal that the method of structured light model and simulation proposed are reasonable and effective, which have important theoretical value and practical signification.

Experiments of depth detection and image recovery of a remote target using a complex hologram

Depth detection and image recovery of a target from its complex hologram have recently been proposed in optical remote sensing applications. The proposed technique involves synthesizing the so-called real-only spectrum hologram from the complex hologram. The resulting spectrum hologram contains the real-only information of the target's spectrum and carries the target's depth information. By digital processing of the spectrum hologram, we can extract the target's depth. Once the depth is obtained, we can recover the image of the target from its complex hologram. We report an experimental verification of the technique by optical acquisition of the complex hologram of the remote target and subsequently extract the target's depth. The image of the target at its depth location is subsequently recovered digitally. The novelty of this remote sensing technique is that by performing a single 2-D scan, we can acquire simultaneously the target's depth information and its image for possible subsequent optical correlation applications.

Timing modulation of a 40-MHz laser-pulse train for target ranging and identification.

We have demonstrated a target ranging and identification technique based on the timing modulation of a mode-locked laser coupled with fast incoherent detection. The range-to-target and the target-depth information have been determined with a resolution of better than 25 cm at single-pulse signal-to-noise ratios below 0.1. Our modeling results suggest that laser average power requirements remain a challenge, with upwards of 100 W likely needed for extension of this technique to ranges over 10 km, but improvements in overall system throughput would allow realization of its potential.

Depth-based target segmentation for intelligent vehicles: fusion of radar and binocular stereo

Dynamic environment interpretation is of special interest for intelligent vehicle systems. It is expected to provide lane information, target depth, and the image positions of targets within given depth ranges. Typical segmentation algorithms cannot solve the problems satisfactorily, especially under the high-speed requirements of a real-time environment. Furthermore, the variation of image positions and sizes of targets creates difficulties for tracking. In this paper, we propose a sensor-fusion method that can make use of coarse target depth information to segment target locations in video images. Coarse depth ranges can be provided by radar systems or by a vision-based algorithm introduced in the paper. The new segmentation method offers more accuracy and robustness while decreasing the computational load.

Design of near-infrared imaging probe with the assistance of ultrasound localization.

A total of 364 optical source-detector pairs were deployed uniformly over a 9 cm x 9 cm probe area initially, and then the total pairs were reduced gradually to 60 in experimental and simulation studies. For each source-detector configuration, three-dimensional (3-D) images of a 1-cm-diameter absorber of different contrasts were reconstructed from the measurements made with a frequency-domain system. The results have shown that more than 160 source-detector pairs are needed to reconstruct the absorption coefficient to within 60% of the true value and appropriate spatial and contrast resolution. However, the error in target depth estimated from 3-D images was more than 1 cm in all source-detector configurations. With the a priori target depth information provided by ultrasound, the accuracy of the reconstructed absorption coefficient was improved by 15% and 30% on average, and the beam width was improved by 24% and 41% on average for high- and low-contrast cases, respectively. The speed of reconstruction was improved by ten times on average.