Survey and Application of Target Detection Algorithms Based on Deep Learning

Tianyu Li ； Hao Wu ； Yanling Mao ； Dong Li ； Lei Chen

《World Scientific Research Journal》 6卷8期 (2020/08) Pp. 76-85

https://doi.org/10.6911/WSRJ.202008_6(8).0010

摘要

This paper reviews the target detection algorithm based on deep learning and its application in the power job site. Firstly, the development of the target detection algorithm and the research directions in recent years are sorted out; then the single-stage and two-stage detection algorithms are separately reviewed. To elaborate, describe its advantages and disadvantages in detail, and on this basis, analyze in detail the improved network architecture based on two methods, and at the same time analyze the application of target detection technology in the power job site; finally, the current stage of deep learning target detection The shortcomings of the algorithm and its development direction are discussed.

關鍵字

Deep learning ； Target detection ； One- stage ； Two-stage ； Power job site

參考文獻

Ríha Kamil, Mašek Jan, Burget Radim, Beneš Radek, Závodná Eva. Novel method for localization of common carotid artery transverse section in ultrasound images using modified Viola-Jones detector. [J]. Ultrasound in medicine & biology,2013,39(10).

Hou Beiping, Zhu Wen. Fast Human Detection Using Motion Detection and Histogram of Oriented Gradients. 2011, 6(8):1597-1604.

M. Fatih Talu, Mehmet Gül, Nuh Alpaslan, Birgül Yiğitcan. Calculation of melatonin and resveratrol effects on steatosis hepatis using soft computing methods[J]. Computer Methods and Programs in Biomedicine,2013,111(2).

Girshick R, Donahue J, Darrell and T, etal. Rich feature hierarchies for object detection and semantic segmentation[C]//2014 IEEE Conf. on Computer Vision and Pattern Recognition, 2014;1-21.

He K, Zhang X, Ren S, etal. Spatial pyramid pooling in deep convolutional networks for visual recognition[J]. IEEE Trans. onPattern Analysis & MachineIntelligence, 2014, 37(9): 1904-1916.