Efficient visual tracking via sparse representation and back-projection histogram

Sliti, Oumaima; Hamam, Habib

doi:10.1007/s11042-019-7439-1

Cited by 6 publications

(2 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The Siamese neural network model needs to be run twice to get the loss. For video TT, the distance between the template frame and the current frame is measured [18].…”

Section: Tt Algorithm Based On Siamese Cnnmentioning

confidence: 99%

Algorithm Research of Neural Network for Feature Target Tracking

2022

View full text Add to dashboard Cite

In recent years, with the explosive growth of research in the field of computer vision, TT has an important application in visual recognition tasks. It can assist target detection and improve the speed of recognition, which has certain theoretical value and research significance. However, in practical application scenarios, target tracking (TT) still faces the problems of inaccurate tracking, poor robustness, and low overall system speed caused by scene changes. Since the TT algorithm based on convolutional neural network (CNN) was proposed, it has attracted the attention of a large number of researchers with the advantages of both speed and accuracy. The multi-layer convolution features extracted from the input image through the CNN have a good appearance representation ability for the target under a variety of complex interference factors. Therefore, this paper uses the CNN to build a feature TT algorithm model, and experiments prove that the accuracy of the model can be improved by increasing the order of magnitude and deepening the ResNet backbone network design. After comparing the tracking effects of the CNN algorithm and other algorithms, it shows that the CNN algorithm has the highest tracking accuracy and success rate.

show abstract

“…The Siamese neural network model needs to be run twice to get the loss. For video TT, the distance between the template frame and the current frame is measured [18].…”

Section: Tt Algorithm Based On Siamese Cnnmentioning

confidence: 99%

Algorithm Research of Neural Network for Feature Target Tracking

2022

View full text Add to dashboard Cite

show abstract

“…There are statistical-based methods, model-based methods, structure-based methods and signal processing-based methods for extracting texture features. The gray level co-occurrence matrix based on statistical methods is used to extract the texture feature because of its ability to resist noise, simple extraction, fast processing speed (Wu et al, 2017;Sliti and Hamam, 2019).…”

Section: Texture Feature Extractionmentioning

confidence: 99%

An improved particle filtering to locate the crop boundary of an unharvested region using vision

Wang

Qin

et al. 2020

View full text Add to dashboard Cite

Purpose Accurately, positioning is a fundamental requirement for vision measurement systems. The calculation of the harvesting width can not only help farmers adjust the direction of the intelligent harvesting robot in time but also provide data support for future unmanned vehicles. Design/methodology/approach To make the length of each pixel equal, the image is restored to the aerial view in the world coordinate system. To solve the problem of too much calculation caused by too many particles, a certain number of particles are scattered near the crop boundary and the distribution regularities of particles’ weight are analyzed. Based on the analysis, a novel boundary positioning method is presented. In the meantime, to improve the robustness of the algorithm, the back-projection algorithm is also used for boundary positioning. Findings Experiments demonstrate that the proposed method could well meet the precision and real-time requirements with the measurement error within 55 mm. Originality/value In visual target tracking, using particle filtering, a rectangular is used to track the target and cannot obtain the boundary information. This paper studied the distribution of the particle set near the crop boundary and proposed an improved particle filtering algorithm. In the algorithm, a small amount of particles is used to determine the crop boundary and accurate positioning of the crop boundary is realized.

show abstract

Online content-based sequential recommendation considering multimodal contrastive representation and dynamic preferences

Lu,

Duan

2024

Neural Comput & Applic

View full text Add to dashboard Cite

Efficient visual tracking via sparse representation and back-projection histogram

Cited by 6 publications

References 44 publications

Algorithm Research of Neural Network for Feature Target Tracking

Algorithm Research of Neural Network for Feature Target Tracking

An improved particle filtering to locate the crop boundary of an unharvested region using vision

Online content-based sequential recommendation considering multimodal contrastive representation and dynamic preferences

Contact Info

Product

Resources

About