Optimization method for autonomous patrol point location of unmanned aerial vehicle and target state recognition method

The invention relates to the technical field of intelligent inspection of power equipment, in particular to an unmanned aerial vehicle autonomous inspection point location optimization method and a target state recognition method, and the method comprises the steps: constructing image pairs in a to-be-inspected object three-dimensional model according to a reference image and a real image corresponding to each initial inspection point location, and determining each matching matrix; according to each matching matrix, obtaining and according to a rotation angle, a scaling factor and a translation vector corresponding to each initial inspection point location, and sequentially judging whether each initial inspection point location needs to be optimized, so that the inspection point location optimization judgment accuracy is improved; the target inspection point location is obtained by adjusting the size of each parameter corresponding to the current initial inspection point location, so that the deviation betwee