向量相乘时，向量维度不一致

[intelligence66]

函数ed.DoDCTTrans(img,blksize)报错了，我打印了DCTcoeffs.shape: (24, 450)的形状，以及每个block的形状：block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 8)
block.shape: (8, 2)。发现是最后一个block形状不符合blksize数值的大小。针对这个错误，我刚开始对最后一张block做了填充，让他的形状为：（8， 8）。往后运行ed.DoInvDCTTrans_and_DeQuantize(DCTcoeffs,blksize,200*std_luminance_quant_tbl)和ed.DoZigZagScan(DCTcoeffs,blksize,scan_order_row)也报这个错误。针对这个问题，请问有解决办法吗？如有打扰，请见谅