Spatial Statistic
Spatial autocorrelation metrics evaluate whether chemical finish concentration values across a fabric web are clustered, dispersed, or randomly distributed. In textile quality engineering, the Moran index calculates spatial clustering of surface coatings using grid-sampled spectroscopic density measurements. The resulting numerical value identifies systematic finishing defects such as streakiness, pad-roll channeling, or uneven spray nozzle coverage.
The mathematical evaluation applies strictly to spatial spatial-array data grids without measuring total chemical content.
Autocorrelation Formula
Calculations compare localized chemical concentration values at specific spatial coordinates against the global mean value across the entire sample area. Values range from negative one to positive one. A positive score approaching positive one indicates strong spatial clustering, where high chemical concentrations concentrate in continuous bands or spots.
A score near zero indicates complete spatial randomness, reflecting ideal microscopic chemical scattering across the fabric surface. Negative values indicate disperse checkerboard patterns, which rarely occur in liquid finishing processes.
Finish Clumping
Mill quality teams use spatial autocorrelation analysis to troubleshoot continuous dyeing and finishing ranges. High positive Moran values highlight systematic process errors, including unequal nip roller pressure, migrating binder resins during infrared drying, or clogged spray manifolds. Resolving spatial clustering improves localized hydrostatic performance and eliminates visual shade variations across broadloom fabrics.
Mapping Limit
Index accuracy depends directly on spatial sampling grid resolution and total sample size. Coarse sampling grids miss micro-scale finish variations between individual yarns.
