Object detection and depth estimation of real world objects using single camera

Liaquat, Sana; Khan, Umar Shahbaz; ur-Rehman, Ata

doi:10.1109/icase.2015.7489526

Cited by 5 publications

(5 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For computing the motion correction, we must first take into account what is happening in the reference frame of the target vehicle. Since we only get the relative speed of a vehicle from a sensor (V relx , V rely ), we have to compute its absolute speed (V x , V y ) in both longitudinal (x) and lateral (y) directions (1). The ego speed on the x and y axes is denoted by…”

Section: Object Spatial and Temporal Alignmentmentioning

confidence: 99%

“…Such approaches are compelling because the object class can also be retrieved. Solutions based on mono-cameras are more affordable; however, they are unable to accurately detect the distance to objects when the road is not flat [1]. Stereo camera devices can solve this problem; however, a major limitation of such systems is that they cannot reconstruct the environment reliably for multiple reasons such as bad illumination conditions, solar flares, perspective warping effect and lack of texture, among others.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Stabilization and Validation of 3D Object Position Using Multimodal Sensor Fusion and Semantic Segmentation

Muresan

Giosan

Nedevschi

2020

Sensors

View full text Add to dashboard Cite

The stabilization and validation process of the measured position of objects is an important step for high-level perception functions and for the correct processing of sensory data. The goal of this process is to detect and handle inconsistencies between different sensor measurements, which result from the perception system. The aggregation of the detections from different sensors consists in the combination of the sensorial data in one common reference frame for each identified object, leading to the creation of a super-sensor. The result of the data aggregation may end up with errors such as false detections, misplaced object cuboids or an incorrect number of objects in the scene. The stabilization and validation process is focused on mitigating these problems. The current paper proposes four contributions for solving the stabilization and validation task, for autonomous vehicles, using the following sensors: trifocal camera, fisheye camera, long-range RADAR (Radio detection and ranging), and 4-layer and 16-layer LIDARs (Light Detection and Ranging). We propose two original data association methods used in the sensor fusion and tracking processes. The first data association algorithm is created for tracking LIDAR objects and combines multiple appearance and motion features in order to exploit the available information for road objects. The second novel data association algorithm is designed for trifocal camera objects and has the objective of finding measurement correspondences to sensor fused objects such that the super-sensor data are enriched by adding the semantic class information. The implemented trifocal object association solution uses a novel polar association scheme combined with a decision tree to find the best hypothesis–measurement correlations. Another contribution we propose for stabilizing object position and unpredictable behavior of road objects, provided by multiple types of complementary sensors, is the use of a fusion approach based on the Unscented Kalman Filter and a single-layer perceptron. The last novel contribution is related to the validation of the 3D object position, which is solved using a fuzzy logic technique combined with a semantic segmentation image. The proposed algorithms have a real-time performance, achieving a cumulative running time of 90 ms, and have been evaluated using ground truth data extracted from a high-precision GPS (global positioning system) with 2 cm accuracy, obtaining an average error of 0.8 m.

show abstract

Section: Object Spatial and Temporal Alignmentmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Stabilization and Validation of 3D Object Position Using Multimodal Sensor Fusion and Semantic Segmentation

Muresan

Giosan

Nedevschi

2020

Sensors

View full text Add to dashboard Cite

show abstract

“…We employ a depth estimation technique developed by Liaquat et al [15]. The proposed depth estimation technique uses only a single camera.…”

Section: Depth Estimationmentioning

confidence: 99%

A Smart Vision Based Navigation Aid for the Visually Impaired

Kommey

Herrman

Addo

2019

AJRCoS

View full text Add to dashboard Cite

Due to the ever increasing number of blind and visually impaired people in the world, there has been a great amount of research dedicated to the design of assistive technologies to support them. The various assistive technologies apply different techniques including laser, ultrasonic sensors and image processing. Autonomous navigation is a significant challenge for the visually impaired, it makes life uncomfortable for them and poses serious safety issues. In this paper we review the progress made so far in vision based systems and propose an approach for developing navigation aids through techniques used in other autonomous systems like self-driving vehicles. The proposed system uses a front camera to capture images and then produces commensurate guiding audio signals that allow the user freely move in their environment. An extra rear camera is included to allow the user to obtain more information about the scene. Care is taken however not to overload the user with information. The proposed method is tested both in indoor and outdoor scenes and is effective in notifying the user for any obstacles. The goal of this paper is to propose a model for and to develop subsystems for an intelligent, high performance, affordable and easy to use image based navigation aid for the visually impaired.

show abstract

“…In addition to using robots, cameras can also be used to identify objects. In 2015, Liaquat et al (Liaquat et al, 2015) proposed a single camera-based depth estimation technique. After detecting different objects in the room, they calculate their areas, and then estimate the distance between the camera and the object based on the training data and a polynomial curve fitting method.…”

Section: Related Workmentioning

confidence: 99%

Object-Locating via Object Recognition and Indoor Positioning

Yang¹,

Chin²

2019

International Conferences Interfaces and Human Computer Interaction 2019; Game and Entertainment Technologies 2019; And Compute

View full text Add to dashboard Cite

Object locating has been a problem in our daily lives, and some techniques have been proposed to address this issue. However, there are still some drawbacks with previous methods. This paper proposes to use two techniques to make object locating easier through object recognition and indoor positioning, so that not only a picture containing the desired object can be shown to help recall where the object was placed, but also the location or room can be shown to help guide where to retrieve the desired object.

show abstract

Object detection and depth estimation of real world objects using single camera

Cited by 5 publications

References 10 publications

Stabilization and Validation of 3D Object Position Using Multimodal Sensor Fusion and Semantic Segmentation

Stabilization and Validation of 3D Object Position Using Multimodal Sensor Fusion and Semantic Segmentation

A Smart Vision Based Navigation Aid for the Visually Impaired

Object-Locating via Object Recognition and Indoor Positioning

Contact Info

Product

Resources

About