Code
Hub
Workspaces
Following
Trending
Connect
MCP
copy
Create free account
hub
/
github.com/YikiDragon/SolarPanelDefectDetect
/ functions
Functions
61 in github.com/YikiDragon/SolarPanelDefectDetect
⨍
Functions
61
◇
Types & classes
9
↓ 14 callers
Function
trough_clustering
波谷数据聚合,过于接近的序号合并为1个序号, 以左右距离阈值为10进行聚合 :param trough: 波谷序号数组 :param th: 同类点最大距离 :return: trough_clustered 聚合后的波谷
image_utils.py:179
↓ 8 callers
Method
updatePrompt
(self)
main.py:286
↓ 6 callers
Method
update_config
(self, conf_path="config.json")
main.py:270
↓ 5 callers
Function
correct
太阳能板姿态校正 :param image: 待校正图像 :param hist_th: 直方图双峰判断阈值,设在直方图里两波峰之间,实验综合最佳值为25 :param approx_eps: 四边形逼近系数初值,会动态调整 :param debug:
image_utils.py:9
↓ 5 callers
Function
segment
电池板自动化分割 :param image_corrected: 校正后的电池板图 :param trough_th: 波谷判别阈值,用于确定分割位置(弃用) :param seg_method: 分割方法,0-纯波谷分
image_utils.py:124
↓ 5 callers
Method
updateItemList
(self)
main.py:280
↓ 3 callers
Method
check_images
(self)
main.py:274
↓ 3 callers
Function
get_dataset
读数据集中各图片路径与标签 输入:root_dir 根目录文件夹 输出:all_image_paths 各图片的路径 all_image_labels 各图片的标签
DenseNet/utils.py:25
↓ 3 callers
Function
get_dataset
读数据集中各图片路径与标签 输入:root_dir 根目录文件夹 输出:all_image_paths 各图片的路径 all_image_labels 各图片的标签
SVM_Kernel/utils.py:25
↓ 3 callers
Function
load_and_preprocess_image
根据路径读取图片张量并统一大小
DenseNet/utils.py:97
↓ 3 callers
Function
load_and_preprocess_image
根据路径读取图片张量并统一大小
SVM_Kernel/utils.py:97
↓ 3 callers
Function
loss_func
(Model: SVMModel, C)
SVM_Kernel/KernelSVM_model.py:62
↓ 2 callers
Method
__init__
(self, bottleneck_num, growth_rate)
DenseNet/DenseNet.py:5
↓ 2 callers
Function
fft_to_rc
(wave, r=100)
image_utils.py:430
↓ 1 callers
Method
Generate_PR
(self)
DenseNet/utils.py:193
↓ 1 callers
Method
Generate_PR
(self)
SVM_Kernel/utils.py:193
↓ 1 callers
Method
Record_output
对单张图片记录输出 :param true_labels: 真实标签 [n,1,1] :param pred_scores: 预测分数 [n,m,c] n个位置可能产生m个不同形状的预测框,每个框有c个预测类别
DenseNet/utils.py:129
↓ 1 callers
Method
Record_output
对单张图片记录输出 :param true_labels: 真实标签 [n,1,1] :param pred_scores: 预测分数 [n,m,c] n个位置可能产生m个不同形状的预测框,每个框有c个预测类别
SVM_Kernel/utils.py:129
↓ 1 callers
Method
__init__
(self, saved_X, saved_Y, kernel)
SVM_Kernel/KernelSVM_model.py:7
↓ 1 callers
Function
balanced_dataset
:param data_list: 数据标签列表 :param sample_refer: 取样参考,0参考损坏集,1参考完好集 :param sample_rate: 取样率 :param label_ratio: 标签比
DenseNet/utils.py:50
↓ 1 callers
Function
balanced_dataset
:param data_list: 数据标签列表 :param sample_refer: 取样参考,0参考损坏集,1参考完好集 :param sample_rate: 取样率 :param label_ratio: 标签比
SVM_Kernel/utils.py:50
↓ 1 callers
Method
compute_AP
(self)
DenseNet/utils.py:214
↓ 1 callers
Method
compute_AP
(self)
SVM_Kernel/utils.py:214
↓ 1 callers
Function
get_dataset
读数据集中各图片路径与标签 输入:root_dir 根目录文件夹 输出:all_image_paths 各图片的路径 all_image_labels 各图片的标签
label_convert.py:5
↓ 1 callers
Method
judgment
(self, result, model='SVM')
main.py:309
↓ 1 callers
Method
plot_PR_curve
(self, ClassNameList, title, fig_path=None)
DenseNet/utils.py:235
↓ 1 callers
Method
plot_PR_curve
(self, ClassNameList, title, fig_path=None)
SVM_Kernel/utils.py:235
↓ 1 callers
Method
read_config
(self, conf_path="config.json")
main.py:235
↓ 1 callers
Method
segArrayPreprocess
(self, image: np.ndarray, model='SVM')
main.py:290
Method
IoU
计算IoU :param bbox1: [4,] Tensor :param bbox2: [4,] Tensor :return IoU:
DenseNet/utils.py:139
Method
IoU
计算IoU :param bbox1: [4,] Tensor :param bbox2: [4,] Tensor :return IoU:
SVM_Kernel/utils.py:139
Method
__init__
(self)
main.py:21
Method
__init__
(self, out_channel)
DenseNet/DenseNet.py:44
Method
__init__
(self, class_num)
DenseNet/DenseNet.py:58
Method
__init__
(self, class_num=7)
DenseNet/utils.py:114
Method
__init__
(self)
DenseNet/model.py:6
Method
__init__
(self, class_num=7)
SVM_Kernel/utils.py:114
Method
__init__
(self, saved_X, saved_Y, kernel)
SVM_Kernel/KernelSVM_model.py:37
Method
build
(self, inputshape)
SVM_Kernel/KernelSVM_model.py:24
Method
call
(self, input_tensor, training=False)
DenseNet/DenseNet.py:18
Method
call
(self, input_tensor, training=False)
DenseNet/DenseNet.py:51
Method
call
(self, input_tensor, training=False)
DenseNet/DenseNet.py:64
Method
call
(self, input_tensor)
DenseNet/model.py:28
Method
call
(self, inputs)
SVM_Kernel/KernelSVM_model.py:31
Method
call
(self, inputs)
SVM_Kernel/KernelSVM_model.py:41
Method
default
(self, line)
main.py:203
Method
do_about
Show author information
main.py:194
Method
do_detect
Detection command: start detection after setting pictures and models
main.py:138
Method
do_exit
Exit solaraneldefectdetect v1.0
main.py:133
Method
do_set
Set the necessary parameters.\nExample: set [options] <param>
main.py:48
Method
do_show
Usage:\n show <command>\n\nDescription:\n Show information or options available for the current operation.\n
main.py:29
Method
emptyline
(self)
main.py:207
Function
exchange_Y_u
判断是否有可能错误的两个伪标记,并将他们交换 :param y_u_pred: # 预测值 :param Y_u: # 伪标记 :return: y_u_pred # 原预测值 Y_u # 交换后的伪
SVM_Kernel/KernelSVM_model.py:111
Function
gamma_norm
伽马归一化 :param image: 图片 :param gamma: 伽马值 :return: 处理后图片
DenseNet/utils.py:9
Function
gamma_norm
伽马归一化 :param image: 图片 :param gamma: 伽马值 :return: 处理后图片
SVM_Kernel/utils.py:9
Function
kernel_gaussian
:param x_data: 参与生成的X数据矩阵 :param prediction_grid: 待预测的单个x数据向量 :param gamma: 方差倒数 :return:
SVM_Kernel/KernelSVM_model.py:46
Function
loss_func
(y_true, y_pred)
DenseNet/model.py:41
Function
loss_func_tsvm
:param Model: SVM模型 :param Cl: 全监督权重 :param Cu: 无监督权重 :param y_u_pred: 无监督模型预测值 :param split_Cu: 是否拆分无监督权重 :ret
SVM_Kernel/KernelSVM_model.py:78
Method
preloop
(self)
main.py:211
Method
reset_state
(self)
DenseNet/utils.py:262
Method
reset_state
(self)
SVM_Kernel/utils.py:262