Development of a lossless compression method applicable to orthophotos and point cloud data

2023 Fiscal Year Final Research Report

• Project/Area Number: 21K11972
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Review Section: Basic Section 61010:Perceptual information processing-related
• Research Institution: Shizuoka Institute of Science and Technology (2023); Maizuru National College of Technology (2021-2022)
• Principal Investigator: Ashizawa Keita 静岡理工科大学, 情報学部, 教授 (70548073)
• Project Period (FY): 2021-04-01 – 2024-03-31
• Keywords: 画像処理 / 情報圧縮 / 可逆圧縮 / 連長符号化

Outline of Final Research Achievements

We studied on information compression of new types of images that we see more and more in our daily life, such as aerial photography by drones, point cloud data obtained by surveying, and ultra-high-resolution images. At first, we looked for statistical properties common to each image signal, but we could not find them. We then shifted our research focus to the coding part, which is a common process. By looking at the signal sequence from a different perspective than before, we succeeded in developing a new method, which we named slice run-length. As a result, we were able to confirm a significant information compression effect for all test images in which we conducted experiments, compared to a method using the Runlength coding.

Free Research Field

知覚情報処理

Academic Significance and Societal Importance of the Research Achievements

従来の視覚的に不必要な成分を取捨選択する非可逆圧縮方式は、画像の品質とデータ量のトレードオフが基本となる。しかし高品質な画像に対し、圧縮のためとはいえ情報を捨てることは本末転倒な状況といえる。そこで本申請課題では、空撮画像や点群データなどの新たに登場した画像と、何も情報を捨てない可逆方式による画像圧縮に注目した。本課題において、画像の信号解析を丁寧に行ったことで、符号化部における連長符号化を置き換える方式のアイデアを得た。スライス連長符号と名付けた提案アルゴリズムは、既存の画像圧縮方式に組み入れることも可能であるため、多くの方式の効率向上に寄与できるものと考えている。