All solutions are designed for fully automatic, large-scale Earth observation workflows and seamless integration into operational processing pipelines.

Products & Solutions

KSKY Hi-Tech delivers new-generation, fully automatic technologies for optical satellite and aerial imagery. Our solutions are designed to improve image clarity, radiometric reliability, and information usability for large-scale Earth observation, environmental monitoring, and intelligent analytics.

AICOOL – Fully Automatic Haze Removal & Image Clarification

AICOOL is a new-generation, fully automatic haze removal and image clarification technology for high-resolution optical satellite and aerial imagery. Unlike conventional atmospheric correction approaches, AICOOL adopts an image-driven framework based on a novel electromagnetic wave component formulation, enabling the separation of haze, thin clouds, and surface information without relying on sensor metadata or external atmospheric inputs. The technology is specifically designed for large-scale, high-throughput batch processing and performs robustly under complex and spatially heterogeneous haze conditions, while preserving radiometric and spectral consistency required for quantitative analysis.

    Key Features

  • New-generation, image-driven haze removal fundamentally different from radiative-transfer-based methods
  • Fully automatic processing without sensor metadata or auxiliary atmospheric information
  • High-speed batch processing for large image archives
  • Robust performance under non-uniform haze and thin cloud conditions
  • Suitable for ground target recognition, land cover mapping, and change detection

AI-Based PM2.5 Estimation from Satellite Imagery

This solution provides fully automatic estimation of PM2.5 air pollution using high-resolution optical satellite imagery. It is built upon haze distribution extraction derived from AICOOL processing and integrates AI-driven modeling and quality control to support large-scale operational monitoring. The system enables county-level or district-level PM2.5 assessment and is suitable for the deployment of automated satellite-based air-quality monitoring stations.

    Key Features

  • High-accuracy PM2.5 estimation based on image-derived haze information
  • Fully automated processing with integrated quality control
  • Scalable from local to regional monitoring applications
  • Compatible with Sentinel, Gaofen, ZY-3 (Resource-3), and other optical sensors
  • Designed for environmental agencies and air-quality research

Multi-Satellite Image Radiometric Normalization

This solution provides a robust framework for radiometric normalization and harmonization across multiple satellite platforms. It is based on the author’s pioneering research in radiometric normalization, compositing, and large-area image mosaicking, and ensures consistent reflectance behavior across sensors, acquisition times, and imaging conditions. The system supports stable long-term time series analysis and improves the reliability of downstream analytics, including AI-based applications.

    Key Features

  • Cross-sensor radiometric consistency for multi-satellite analysis
  • Essential for large-area mosaics and long-term environmental monitoring
  • Improves accuracy and stability of downstream AI and analytical workflows
  • Suitable for both commercial and governmental Earth observation missions