|
|
|
|
|
import numpy as np
|
|
|
|
|
|
import lqmtest_x3_2_load_data
|
|
|
|
|
|
|
|
|
|
|
|
class ModelObj: # 网络对象
|
|
|
|
|
|
def __init__(self, ObjID, ObjType, ObjLable, ParaString, ObjX, ObjY):
|
|
|
|
|
|
self.ObjID = ObjID # 图元号
|
|
|
|
|
|
self.ObjType = ObjType # 图元类别
|
|
|
|
|
|
self.ObjLable = ObjLable # 对象标签
|
|
|
|
|
|
self.ParaString = ParaString # 参数字符串
|
|
|
|
|
|
self.ObjX = ObjX # 对象位置x坐标
|
|
|
|
|
|
self.ObjY = ObjY # 对象位置y坐标
|
|
|
|
|
|
class Conv_Class(ModelObj): # 卷积对象
|
|
|
|
|
|
def __init__(self, ObjID, ObjType, ObjLable, ParaString, ObjX, ObjY):
|
|
|
|
|
|
super().__init__(ObjID, ObjType, ObjLable, ParaString, ObjX, ObjY)
|
|
|
|
|
|
self.ConvProc = self.conv_proc # 基本操作函数
|
|
|
|
|
|
self.SetConvPara = self.setconv_para # 参数设置函数
|
|
|
|
|
|
# 定义卷积函数ConvProc()
|
|
|
|
|
|
def conv_proc(self, image, ConvPara):
|
|
|
|
|
|
# 获取输入数据的大小，这里假设是单通道的图片
|
|
|
|
|
|
c, image_h, image_w = image.shape
|
|
|
|
|
|
kernel_h = ConvPara["kernel_h"]
|
|
|
|
|
|
kernel_w = ConvPara["kernel_w"]
|
|
|
|
|
|
kernel = ConvPara["kernel"]
|
|
|
|
|
|
out_h = (image_h - kernel_h) // ConvPara["stride"] + 1
|
|
|
|
|
|
out_w = (image_w - kernel_w) // ConvPara["stride"] + 1
|
|
|
|
|
|
output = np.zeros((c, out_h, out_w)) # 初始化输出数据为零矩阵
|
|
|
|
|
|
for k in range(c): # 遍历每个通道
|
|
|
|
|
|
for i in range(out_h): # 遍历每个输出位置
|
|
|
|
|
|
for j in range(out_w):
|
|
|
|
|
|
stride = ConvPara["stride"] # 获得步长
|
|
|
|
|
|
output[k, i, j] = np.sum(
|
|
|
|
|
|
image[k, i * stride:i * stride + 3, j *
|
|
|
|
|
|
stride:j * stride + 3] * kernel)#计算卷积
|
|
|
|
|
|
return output # 返回卷积计算后的数组（特征向量）
|
|
|
|
|
|
def setconv_para(self): # 定义设置卷积参数的函数SetConvPara()
|
|
|
|
|
|
kernel_h = int(input("请输入卷积核的高度: ")) # 用户输入卷积核的高度 3
|
|
|
|
|
|
kernel_w = int(input("请输入卷积核的宽度: ")) # 用户输入卷积核的宽度 3
|
|
|
|
|
|
# 用户输入卷积核的值
|
|
|
|
|
|
kernel = []
|
|
|
|
|
|
# [[1.289202, -1.471377, -0.238452],
|
|
|
|
|
|
# [-0.562343, -0.019988, -0.441446],
|
|
|
|
|
|
# [1.627381, 1.390266, 0.812486]]
|
|
|
|
|
|
print("请输入卷积核的值（一行一行输入，各值之间用空格分开）：")
|
|
|
|
|
|
for i in range(kernel_h):
|
|
|
|
|
|
row = [float(val) for val in input().split()]
|
|
|
|
|
|
kernel.append(row)
|
|
|
|
|
|
|
|
|
|
|
|
stride = int(input("请输入步长: ")) # 用户输入步长 1
|
|
|
|
|
|
padding = int(input("请输入填充: ")) # 用户输入填充 0
|
|
|
|
|
|
|
|
|
|
|
|
# 返回ConvPara参数，这里用一个字典来存储
|
|
|
|
|
|
ConvPara = {"kernel": kernel, "kernel_h": kernel_h,
|
|
|
|
|
|
"kernel_w": kernel_w, "stride": stride,
|
|
|
|
|
|
"padding": padding}
|
|
|
|
|
|
return ConvPara
|
|
|
|
|
|
if __name__ == '__main__':
|
|
|
|
|
|
|
|
|
|
|
|
DataSet = lqmtest_x3_2_load_data.Data_Class("DataSet1", 1, "数据集1", [], 120, 330)
|
|
|
|
|
|
# setload_data()函数，获取加载数据集的参数
|
|
|
|
|
|
DataPara = DataSet.SetDataPara()
|
|
|
|
|
|
train_images, test_images = DataSet.LoadData(DataPara)
|
|
|
|
|
|
|
|
|
|
|
|
Conv = Conv_Class("Conv1", 2, "卷积1", [], 250, 330)
|
|
|
|
|
|
ConvPara = Conv.SetConvPara()
|
|
|
|
|
|
for i in range(len(train_images) // 32):
|
|
|
|
|
|
images = train_images[i * 32:(i + 1) * 32]
|
|
|
|
|
|
conv_images = []# 存储卷积处理后的图片的列表
|
|
|
|
|
|
for image in images:# 获取训练集的图片数据
|
|
|
|
|
|
dim = len(image.shape)# 获取矩阵的维度
|
|
|
|
|
|
if dim == 2: # 如果是二维矩阵,则转化为三维矩阵
|
|
|
|
|
|
image_h, image_w = image.shape
|
|
|
|
|
|
image = np.reshape(image, (1,image_h,image_w))
|
|
|
|
|
|
# 调用ConvProc()函数，根据ConvPara参数完成卷积计算
|
|
|
|
|
|
output = Conv.ConvProc(image, ConvPara)
|
|
|
|
|
|
conv_images.append(output)# 将卷积结果存储到列表
|
|
|
|
|
|
elif dim == 3: # 若为三维矩阵，则保持不变直接卷积处理
|
|
|
|
|
|
output = Conv.ConvProc(image, ConvPara)
|
|
|
|
|
|
conv_images.append(output)
|
|
|
|
|
|
# 将卷积处理后的图片列表转换为数组形式，方便后续处理
|
|
|
|
|
|
conv_images = np.array(conv_images)
|
