A New Approach to Image Compression in Industrial Internet of Things

TL;DR

This paper introduces a novel adaptive lossy image compression method for IIoT sensor images using K-Means++ and IEC, achieving higher quality compression without increasing bandwidth.

Contribution

The study presents the first application of a reusable, color-based clustering technique combining K-Means++ and IEC for improved image compression in IIoT sensor networks.

Findings

Higher image quality at the same compression rate compared to existing clustering methods.

Effective compression of greenhouse sensor images with improved efficiency.

Demonstrated success on real-world dataset from a smart greenhouse.

Abstract

Applying image sensors in automation of Industrial Internet of Things (IIoT) technology is on the rise, day by day. In such companies, a large number of high volume images are transmitted at any moment; therefore, a significant challenge is reducing the amount of transmitted information and consequently bandwidth without reducing the quality of images. Image compression in sensors, in this regard, will save bandwidth and speed up data transmitting. There are several pieces of research in image compression for sensor networks, but, according to the nature of image transfer in IIoT, there is no study in this particular field. In this paper, it is for the first time that a new reusable technique to improve productivity in image compression is introduced and applied. To do this, a new adaptive lossy compression technique to compact sensor-generated images in IIoT by using K- Means++ and Intelligent Embedded Coding (IEC) as our novel approach, is presented. The new method is based on the colour of pixels so that pixels with the same or nearly the same colours are clustered around a centroid and finally, the colour of the pixels will be encoded. The experiments are based on a reputable image dataset from a real smart greenhouse; i.e. KOMATSUNA. The evaluation results reveal that, with the same compression rate, our approach compresses images with higher quality in comparison with other methods such as K-means, fuzzy C-means and fuzzy C-means++ clustering.