A Multiple Feature-Based Image-Switching Strategy in Visual Sensor Networks

Abstract

Generally, one fixed camera is used to take still or dynamic images and extract proper information from the captured images. However, the process of analyzing images through the use of one camera is very sensitive to neighboring environmental factors, such as illumination, background, and noise; thus, it is hard to guarantee precision. To extract proper information from images more precisely in visual sensor networks, this paper proposes an image-switching strategy where, among different types of installed cameras, the one camera best suited to neighboring circumstances is chosen. The proposed strategy is to first receive initial images as input data and then extract multiple features representing neighboring circumstances from the input images. Subsequently, it is to define the neighboring circumstances metric, which is the weighted sum of the extracted features, and to dynamically switch cameras to obtain images in accordance with the neighboring circumstances. The results of the experiment show that the proposed dynamic switching strategy reliably chooses, from among different cameras, the one camera that is best suited to the neighboring circumstances.

A new clothing image retrieval algorithm based on sketch component segmentation in mobile visual sensors

“…Currently, various types of mobile visual sensors such as cameras, mobile phones, and drones have been released and applied to many different areas. 1,2,[11][12][13][14][15] A huge number of different kinds of images have been collected by these sensors. Thus, how to assist user retrieve relevant images for some domain-specific applications is urgently demanded.…”

“…Wang et al 36 proposed convolutional neural network to segment images. Jang and Kim 11 proposed a strategy to extract multiple features representing neighboring circumstances from the input images. Shi et al 3 presented an affine invariant method to produce dense correspondences between uncalibrated wide baseline images.…”