METHOD FOR CLASSIFYING CROP BY MULTIPLE IMAGE RECOGNITION REFLECTING WEIGHT, AND COMPUTER-READABLE RECORDING MEDIUM FOR EXCUTING THE METHOD, AND CROP CLASSIFYING DEVICE

摘要

Provided are a crop classification method using a probability-based score vector derived from a multi-view image, a recording medium in which a program for performing the method is recorded, and a crop classification apparatus therefor. The apparatus for classifying crops includes: an image input unit for receiving images from a plurality of viewpoints of a target object; an individual discriminator for classifying reference data at each time point of the object using an object detection-based deep learning model, and outputting a plurality of probability-based score vectors for each time point regarding the reference data level of the object; a weight calculation unit for calculating a plurality of weights at each time point from a confusion matrix of the score vector; and a final determination unit that determines a final reference data level based on the plurality of probability-based score vectors output from the individual determination unit and the weights calculated from the weight calculation unit.