Microplastics have been one of the most serious environmental concerned pollutants over recent decades. While these pollutants have detected not only in the marine biota, but also in abiotic sea products such as sea salts, which might pose a threat to food safety. Although efficient methods for the analysis of microplastics in different matrices have been published, in many cases, part of the quantification protocol relies on eye counting, which is inefficient and time consuming. In this study, a new method for the analysis of Polyethylene terephthalate (PET) microplastics in table sea salts was developed. After hydrogen peroxide (H2O2) pretreatment and filtration, the PET particles in vacuum-filtered retentates of table sea salts were imaged by micro-FT-NIR imaging and identified using spectral similarity methods including correlation coefficient mapping (CCM), spectral angle mapping (SAM) and Euclidean distance mapping (EDM) methods. The number of the particles were acquired by a computer based automatic counting method. The results showed that, in most cases, the number of microplastics obtained by automatic counting method were equal to those obtained by eye counting method, which proved that micro-NIR chemical imaging combined with chemometrics methods can be an alternative way to identify and count the microplastics simultaneously. This method for detecting and quantifying PET microplastics in table sea salts could also be applied to other types of polymers in other environmental samples detection.