The designation of basic farmland reserves is a strategic requirement for the sustainable development of agriculture of the Chinese government, and it is also a foundation for ensuring national food security and social stability. The quality of cultivated land is an important evidence for the site of protected area, from the perspective of crops themselves, the quality evaluation system should not only consider external factors, but also take internal factors of crops into account. The absorption and utilization results of external conditions of crops in time series were regarded as innovative indicators to break through the traditional evaluation model, therefore, taking Jiayu County, Hubei Province as an example, vegetation index information was introduced to improve the comprehensive quality evaluation system of cultivated land, and then the weight of each evaluation index was calculated by means of analytic network process (ANP). Then local spatial autocorrelation analysis was used, and comparative analysis of the results between the quality of cultivated land and the Moran’s index in pixel scale was processed. The result showed that the overall quality of cultivated land in Jiayu County was good and it showed a strong positive correlation in spatial distribution, the local indicators of spatial association (LISA) of the whole county and cultivated land were 0.8645 and 0.9916, respectively. Besides, both results of demarcation based on pixel gave priority to the HH type (highhigh), 99.97% of pixels were classified as basic farmland. Finally, the former was superior in terms of comprehensive quality, while the latter had good effect on narrow area of the hilly region. The comprehensive quality evaluation system of cultivated land was improved by introducing the information of crops themselves in the pixel scale, which provided a useful reference for the fine delineation of basic farmland.