Satellite Imaging Tech Advances to Detect Car Colors from 500km Above

A single photograph released by The New York Times on X (formerly Twitter) on May 28 captured global attention. The image showed Iranian Supreme Leader Ayatollah Ali Khamenei's residence engulfed in black smoke following a strike that left the compound in ruins. The photograph of the devastated residence was released before any announcement of Khamenei's death, taken by U.S. aerospace company Airbus using satellite imaging.

Analyzing satellite photographs and videos of ground targets for military operations is no longer the stuff of spy movies. Satellite imagery has long been employed in defense applications, particularly for reconnaissance of specific facilities and regions to gather pre-mission intelligence. The technology has proven valuable for its ability to continuously collect target information without enemy interference.

The 2011 operation "Neptune Spear" that killed Osama bin Laden demonstrated satellite imagery's military value. The U.S. collected 387 high-resolution photographs of bin Laden's compound over the month preceding the raid. This satellite-based preparation proved decisive in approving the mission's execution.

Other nations have also adopted satellite imagery as a primary reconnaissance tool. Following the outbreak of the Russia-Ukraine war, Ukraine crowdfunded surveillance satellites for reconnaissance of Russian military facilities, with the images used to prepare Ukrainian military airstrikes. In October 2024, South Korea's National Intelligence Service analyzed satellite photographs to reveal North Korean special forces deployment to the Russia-Ukraine conflict.

Once the exclusive domain of government agencies, satellite imagery is increasingly being commercialized and used for academic research. Small and medium-sized venture companies now launch their own satellites to capture and analyze images. Domestic startups have developed satellite image analysis technology capable of detecting vehicle types and colors from hundreds of kilometers above Earth's surface.

The primary advantage of satellite-based ground imaging is the ability to repeatedly collect data from areas difficult for humans to access. This capability has driven adoption in fields requiring time-series tracking of changes across vast areas, with environmental monitoring and disaster response being prime examples.

"Processing satellite imagery remains challenging, so it's still primarily used in defense," said Cho Sung-ik, CEO of satellite image analysis platform developer Telepix. "However, as technology advances and barriers to entry fall, more industries will require satellite imagery."

All-Weather Imaging and Calibration Technology Determine Image Quality

Key technical factors determining satellite image quality include ultra-high-resolution optical payload development, satellite attitude control, and image calibration and correction. Ultra-high-resolution optical payloads refer to imaging equipment capable of capturing high-quality photographs and video—the adage that good cameras produce good photos applies to satellite imagery as well. Industry experts agree that securing technology across the entire imaging equipment manufacturing process—from optical system design to precision component machining and camera assembly—is essential for obtaining quality images.

The additional challenge lies in creating equipment that operates normally in space. "Space conditions differ vastly from Earth—intense solar radiation, extreme cold," said Kim Sung-hee, CTO of Telepix. "The engineering capability to simulate these extreme environments and incorporate them into design determines image quality."

Satellite attitude control and maneuvering technology is essential for capturing the correct subject within the optical payload's field of view. Photographing desired subjects from Earth orbit requires precise satellite attitude adjustments to aim the field of view. Since satellites photograph Earth from altitudes ranging from several hundred to over 30,000 kilometers, even slight deviations in camera angle result in significant targeting errors, capturing unintended areas.

"At 500 kilometers altitude, a one-degree deviation in camera pointing angle creates approximately 10 kilometers of positional error on the ground," explained Kim Hong-bong, Executive Director at Hancom InSpace. "The ability to accurately and repeatedly photograph customer-designated locations is the core competitive advantage in the satellite imagery industry."

Post-capture calibration and correction also significantly affect image quality. Satellite photography operates under different environmental conditions than ground-based photography. Atmospheric effects, Earth's curvature, and terrain cause various electromagnetic waves to scatter or reflect, creating image distortion. Additionally, satellite viewing angles can introduce tilt in captured subjects. Detecting and correcting all such distortions is necessary for extracting accurate data from satellite imagery.

AI Lowers Barriers to Satellite Image Usage

Recent efforts increasingly integrate artificial intelligence into satellite image analysis, lowering barriers to satellite imagery utilization. Previously, humans had to manually analyze satellite images one by one to extract meaningful information. Now, AI can identify objects within images and extract information based on user requests. "While finding what exists in an image was once the priority, AI now understands context and automatically extracts changes in subjects," Kim noted.

Domestic companies are adopting AI assistants (agents), a trending topic in the IT industry. Users can input commands to AI agents instead of memorizing complex functions in satellite image analysis programs, with AI locating appropriate features and performing some analysis. Telepix launched its AI chatbot service "SatChat" in December 2024. Hancom InSpace has also integrated AI agent functionality into its satellite image analysis platform "InStation." SIA has independently developed AI models for super-resolution (image resolution enhancement) and converting synthetic aperture radar (SAR) images to optical images.

"We've shifted from humans manually reviewing every satellite photograph to AI extracting anomalies while analysts make final judgments," said Jeon Tae-kyun, CEO of SIA. "This AI innovation is essential for processing satellite images that increase by thousands daily."

AI integration is expected to expand the spectrum of civilian satellite imagery applications. Industry observers predict increased satellite imagery usage in industries with high geographic or natural resource dependencies, specifically citing climate change response, agriculture and fisheries, energy resources, real estate, and logistics.

"Future competitiveness in the satellite industry won't simply be about acquiring images," Kim predicted. "The core competitive advantage will be the capability to extract meaningful information faster and more accurately to support user decision-making."