A New Approach To Estimate Depth Of Cars Using A Monocular Image

Tousi, Seyed Mohamad Ali; Khorramdel, Javad; Lotfi, Faraz; Nikoofard, Amirhossein; Ardekani, Ali Najafi; Taghirad, Hamid D.

doi:10.1109/cfis49607.2020.9238702

Cited by 3 publications

(9 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We also evaluated the average loss in comparison to YOLOv3, a similar model used in [28]. YOLOv3 achieved an average training loss of 1.22, while our approach achieved higher accuracy in detecting objects of interest in AVs.…”

Section: Resultsmentioning

confidence: 99%

“…The achieved average loss takes into account the high variance in the size of detected classes (large vehicles compared to small car lights) and the high dynamics of the scenes. We also evaluated the average loss in comparison to YOLOv3, a similar model used in [28]. YOLOv3 achieved an average training loss of 1.22, while our approach achieved higher accuracy in detecting objects of interest in AVs.…”

Section: Resultsmentioning

confidence: 99%

“…This is demonstrated in Figure 7, which shows the implementation of the YOLOv7 algorithm in real-world images, indicating the effectiveness of our provided dataset in achieving the required results. We also evaluated the average loss in comparison to YOLOv3, a similar model used in [28]. YOLOv3 achieved an average training loss of 1.22, while our approach achieved higher accuracy in detecting objects of interest in AVs.…”

Section: Resultsmentioning

confidence: 99%

“…There has been an increasing amount of research on estimating distance using a monocular image in recent years, especially in the areas of inter-vehicle distance estimation [27] and advanced driver assistance systems (ADASs) [5]. Some studies have concentrated on using the bounding box's height to form the object's feature vector, but Tousi et al [28] argued that relying solely on the height of a car in an image may lead to more errors. To reduce sensitivity to errors caused by the placement of bounding boxes, they proposed an alternative approach that is less susceptible to such errors.…”

Section: Introductionmentioning

confidence: 99%

“…The car's angle relative to the center line of the image is indicated by the distance between the line connecting the lights and the center of the detection bounding box. This parameter was previously discussed in [28] as a crucial factor in determining the car's distance from the center of the image in both width and length dimensions.…”

mentioning

confidence: 99%

See 4 more Smart Citations

An Efficient Approach to Monocular Depth Estimation for Autonomous Vehicle Perception Systems

Afshar,

Shirmohammadi,

Ghahramani

et al. 2023

Sustainability

View full text Add to dashboard Cite

Depth estimation is critical for autonomous vehicles (AVs) to perceive their surrounding environment. However, the majority of current approaches rely on costly sensors, making wide-scale deployment or integration with present-day transportation difficult. This issue highlights the camera as the most affordable and readily available sensor for AVs. To overcome this limitation, this paper uses monocular depth estimation as a low-cost, data-driven strategy for approximating depth from an RGB image. To achieve low complexity, we approximate the distance of vehicles within the frontal view in two stages: firstly, the YOLOv7 algorithm is utilized to detect vehicles and their front and rear lights; secondly, a nonlinear model maps this detection to the corresponding radial depth information. It is also demonstrated how the attention mechanism can be used to enhance detection precision. Our simulation results show an excellent blend of accuracy and speed, with the mean squared error converging to 0.1. The results of defined distance metrics on the KITTI dataset show that our approach is highly competitive with existing models and outperforms current state-of-the-art approaches that only use the detected vehicle’s height to determine depth.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Resultsmentioning

confidence: 99%

Section: Resultsmentioning

confidence: 99%