Quality Inspection and Intelligent Fusion Method for Automated Production of Large-Scale Remote Sensing Image Tiles

Abstract. Web map services (such as Google Maps and Amap) have become an indispensable spatial information infrastructure in various fields of society. Their underlying image tile pyramid technology encompasses the entire process of slicing, publishing, and updating massive remote sensing images. In the tile production process, quality inspection and edge blending are crucial steps to ensure the visual authenticity and usability of the final map product. However, current tile production relies on manual sampling for quality inspection, and the blending process suffers from issues such as geometric misalignment and color inconsistency, which hinder production efficiency and product quality. This paper proposes a quality inspection and intelligent blending method for large-scale tile automation production, constructing a four-layer post-processing framework including a metadata quality inspection layer, an image quality inspection layer, an intelligent color coordination layer, and an automatic mosaicking layer. By establishing a rule engine to implement batch verification of metadata, combining pixel statistics and edge detection to achieve precise positioning of black/white edges and color anomaly recognition, using histogram matching and batch color correction to ensure color consistency across multiple batches of tiles, and implementing seamless blending between new tiles and online tiles based on an alpha channel null value filling mechanism, experiments conducted on the TianMap tile dataset show that this method significantly enhances the automation level and processing efficiency of tile production, reduces human errors, and enhances the visual consistency and reliability of web map products. This study provides a technical reference for the intelligent production of remote sensing image tiles and has practical application value in supporting high-quality updates for web map services.