Chain code-based Image Compression and Transfer Schemes for Electronic Shelf Label Systems
- 주제(키워드) Electronic Shelf Label , Chain coding , Image compression , Rateless Code
- 주제(DDC) 621.39
- 발행기관 아주대학교
- 지도교수 김기형
- 발행년도 2022
- 학위수여년월 2022. 2
- 학위명 박사
- 학과 및 전공 일반대학원 컴퓨터공학과
- 실제URI http://www.dcollection.net/handler/ajou/000000031338
- 본문언어 영어
- 저작권 아주대학교 논문은 저작권에 의해 보호받습니다.
초록/요약
One of the most critical and recurring tasks in managing a store is to provide accurate, up-to-date price information to customers on the shelves. Manual updating of price tags has been a time-consuming, error-prone task with high labor costs. An electronic shelf label (ESL) system is an excellent solution to solve many of the problems that traditional paper label systems have, and it is also a foundation system for next-generation services in markets. A typical ESL system configuration in a retail store includes thousands of battery-powered ESL tags that are mostly connected wirelessly in a dense indoor environment. Raising the success ratio of wireless communication is essential for the system's viability due to its limited battery life. Most of the ESL traffic is the image data of goods that appear on the tags, and reducing the amount of the data is one of the most effective ways to enhance communication performance and reduce retransmission. In this dissertation, we propose two schemes to improve the ESL system. One is a new image compression algorithm, ECO, based on chain coding that utilizes the characteristics of the ESL images. The other is a novel media-access-control protocol, ESL-MAC, to transfer images to many ESL tags effectively. ESL images have characteristics that suit the purpose. First, they consist of text, barcodes, and simple figures, and text usually accounts for a large portion of them. Second, there are two or three colors used for them, and last, they have a lot of similar parts. The proposed scheme actively utilizes these characteristics. We elaborate on the proposed methods and demonstrate that they are suitable for improving the performance of ESL systems through performance comparison and simulation with other existing algorithms. In performance evaluation, ECO showed higher compression efficiency and faster decompression speed than other algorithms. ECO can improve the ESL system's overall performance by compensating for the shortcomings of low transmission speed ESL networks and low computational capabilities ESL tags. In addition, ESL-MAC showed that it can effectively update many ESL tags with a shorter time and less traffic generation.
more목차
1. Introduction 1
1.1. Background and statement of the problem 1
1.2. Dissertation objective 5
1.3. Dissertation contribution 6
1.4. Dissertation organization 7
2. Image Compression using Chain Coding for Electronic Shelf Label Systems 8
2.1. Introduction 8
2.2. Motivation and related works 11
2.3. Preliminaries 14
2.3.1. Move-To-Front Transform 14
2.3.2. Chain Coding 15
2.3.3. Variable-Length Integer encoding 16
2.4. Proposed algorithm for image compression in ESL system 17
2.4.1. Data Structure 17
2.4.2. Compression 21
2.4.3. Decompression 31
3. Image Transfer Scheme for Electronic Shelf Labels 38
3.1. Introduction 38
3.2. Related works 41
3.2.1. Rateless Codes 41
3.2.2. Polling & Slot-based MAC 43
3.3. Proposed Scheme 45
3.3.1. Ratio channels 45
3.3.2. Superframe Structure 45
3.3.3. Information in beacons 47
3.3.4. Transmitted image data format 48
3.3.5. Network protocol operation 49
4. Performance evaluation 52
4.1. Performance evaluation of ECO 52
4.1.1. Compressed size 53
4.1.2. Compression time 54
4.1.3. Decompression time 54
4.1.4. Influence of the number of image colors 54
4.1.5. Compression ratio 55
4.2. Performance evaluation of ESL-MAC 70
4.2.1. Transfer to all tags 71
4.2.2. Transfer to some tags 72
4.2.3. Energy consumption 76
5. Conclusions and further studies 78
References 81
