From b37e064594e37916b44dc788f4c250200b29d50c Mon Sep 17 00:00:00 2001 From: FEIJINTI <83849113+FEIJINTI@users.noreply.github.com> Date: Thu, 16 Mar 2023 15:48:03 +0800 Subject: [PATCH] =?UTF-8?q?=E4=BB=A3=E7=A0=81=E6=B5=8B=E8=AF=95=E6=AF=94?= =?UTF-8?q?=E8=BE=83?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- classifer.py | 54 +++++++++++++++++++++++++++++++++++++--------------- hist.py | 22 +++++++++++++++++---- 2 files changed, 57 insertions(+), 19 deletions(-) diff --git a/classifer.py b/classifer.py index 7b3473d..b0de65c 100755 --- a/classifer.py +++ b/classifer.py @@ -14,7 +14,9 @@ import cv2 from sklearn.cluster import KMeans from sklearn.linear_model import LogisticRegression from sklearn.model_selection import train_test_split -from sklearn.metrics import accuracy_score +from sklearn.metrics import accuracy_score, confusion_matrix + + from sklearn.tree import DecisionTreeClassifier from sklearn.neighbors import KNeighborsClassifier from scipy.stats import binom @@ -29,9 +31,9 @@ sys.path.append(os.getcwd()) from root_dir import ROOT_DIR import utils -FEATURE_INDEX = [0, 1, 2] +FEATURE_INDEX = [0,1,2] delete_columns = 10 # 已弃用 -num_bins = 7 +num_bins = 10 class WoodClass(object): @@ -58,9 +60,10 @@ class WoodClass(object): self._single_pick = single_pick_mode self.set_purity(self.pur) self.change_pick_mode(single_pick_mode) - self.model = LogisticRegression(C=1e5) + # self.model = LogisticRegression(C=1e5) self.left_correct = left_correct # self.model = KNeighborsClassifier() + self.model = DecisionTreeClassifier() else: self.load(load_from) self.isCorrect = False @@ -125,25 +128,32 @@ class WoodClass(object): y_pred = self.model.predict(x_test) #将y_pred中1和2转化为1 - y_pred[y_pred == 2] = 1 - y_test[y_test == 2] = 1 + # y_pred[y_pred == 2] = 1 + # y_test[y_test == 2] = 1 + + print(confusion_matrix(y_test, y_pred)) pre_score = accuracy_score(y_test, y_pred) self.log.log("Test accuracy is:" + str(pre_score * 100) + "%.") y_pred = self.model.predict(x_train) - y_pred[y_pred == 2] = 1 - y_train[y_train == 2] = 1 + # y_pred[y_pred == 2] = 1 + # y_train[y_train == 2] = 1 pre_score = accuracy_score(y_train, y_pred) self.log.log("Train accuracy is:" + str(pre_score * 100) + "%.") y_pred = self.model.predict(x) - y[y == 2] = 1 - y_pred[y_pred == 2] = 1 + # y[y == 2] = 1 + # y_pred[y_pred == 2] = 1 pre_score = accuracy_score(y, y_pred) self.log.log("Total accuracy is:" + str(pre_score * 100) + "%.") + + #显示结果报告 + + + return int(pre_score * 100) def calculate_p1(self, x, remove_background=False): @@ -296,11 +306,25 @@ class WoodClass(object): # x = x[np.argsort(x[:, 0])] # x = x[-self.k:, :] + hist, bins = np.histogram(x[:, 0], bins=num_bins) hist = hist[1:] bins = bins[1:] - hist_number = np.argmax(hist) - x = x[(x[:, 0] > bins[hist_number]) & (x[:, 0] < bins[hist_number + 1]), :] + sorted_indices = np.argsort(hist) + hist_number = sorted_indices[-1] + second_hist_number = sorted_indices[-2] + x = x[((x[:, 0] > bins[hist_number]) & (x[:, 0] < bins[hist_number + 1])) | ( + (x[:, 0] > bins[second_hist_number]) & (x[:, 0] < bins[second_hist_number + 1])), :] + + + # hist, bins = np.histogram(x[:, 0], bins=5) + # sorted_indices = np.argsort(hist) + # hist_number = sorted_indices[-1] + # second_hist_number = sorted_indices[-2] + # x = x[((x[:, 0] > bins[hist_number]) & (x[:, 0] < bins[hist_number + 1])) | ( + # (x[:, 0] > bins[second_hist_number]) & (x[:, 0] < bins[second_hist_number + 1])), :] + + if debug_mode: # self.log.log(x) @@ -376,7 +400,7 @@ class WoodClass(object): plt.scatter(x_data[:, 0], x_data[:, 1], c=y_data) plt.show() # 尝试最合适的特征组合，保存提取出的特征的方法 - # 0: l, 1: a, 2: b, 3: var(l), 4: var(a), 5: var(s), 6: h, 7: s, 8: v, 9: var(h) 10: var(s): 11: var(v) + # 0: l, 1: a, 2: b, 3: var(l), 4: var(a), 5: var(b), 6: h, 7: s, 8: v, 9: var(h) 10: var(s): 11: var(v) # 全部：0.941 # [:, [0, 1, 2, 3, 4, 5, 6, 7]] : 0.88 # [:, [0, 1, 2]] : 0.911 @@ -433,7 +457,7 @@ if __name__ == '__main__': settings = Config() # 初始化wood - wood = WoodClass(w=4096, h=1200, n=8000, p1=0.46, debug_mode=False) + wood = WoodClass(w=4096, h=1200, n=8000, p1=0.8, debug_mode=False) print("色彩纯度控制量{}/{}".format(wood.k, wood.n)) data_path = settings.data_path # wood.correct() @@ -442,7 +466,7 @@ if __name__ == '__main__': settings.model_path = str(ROOT_DIR / 'models' / wood.fit_pictures(data_path=data_path)) # 测试单张图片的预测，predict_mode=True表示导入本地的model, False为现场训练的 - pic = cv2.imread(r"data/99/dark/rgb70.png") + pic = cv2.imread(r"data/316/dark/rgb70.png") start_time = time.time() # for i in range(100): wood_color = wood.predict(pic) diff --git a/hist.py b/hist.py index b0a7b75..1ae516b 100644 --- a/hist.py +++ b/hist.py @@ -2,19 +2,33 @@ import cv2 import numpy as np -img_path = 'data/data1103/dark/rgb20.png' +img_path = 'data/99/dark/rgb59.png' img = cv2.imread(img_path) img = cv2.cvtColor(img, cv2.COLOR_BGR2LAB) + +w = img.shape[0] +h = img.shape[1] + +ratio = np.sqrt(5000 / (w * h)) +ww, hh = int(ratio * w), int(ratio * h) +img = cv2.resize(img, (hh, ww)) x = img.reshape(img.shape[0]*img.shape[1], img.shape[2]) + hist, bins = np.histogram(x[:, 0], bins=10) hist = hist[1:] bins = bins[1:] -hist_number = np.argmax(hist) -x = x[(x[:, 0] > bins[hist_number]) & (x[:, 0] < bins[hist_number + 1]), :] + + +sorted_indices = np.argsort(hist) +hist_number = sorted_indices[-1] +second_hist_number = sorted_indices[-2] + +x = x[((x[:, 0] > bins[hist_number]) & (x[:, 0] < bins[hist_number + 1]))|((x[:, 0] > bins[second_hist_number]) & (x[:, 0] < bins[second_hist_number + 1])), :] + mean_value = np.mean(x, axis=0).astype(np.uint8) y = np.zeros((img.shape[0]*img.shape[1], img.shape[2]), dtype=np.uint8) y[:, ] = mean_value #lab转rgb再保存 y = cv2.cvtColor(y.reshape(img.shape[0], img.shape[1], img.shape[2]), cv2.COLOR_LAB2BGR) -cv2.imwrite('5.png', y.reshape(img.shape[0], img.shape[1], img.shape[2])) +cv2.imwrite('59.png', y.reshape(img.shape[0], img.shape[1], img.shape[2]))