From 776a127b8db01cd4338f4db2a84ea567a65bff9f Mon Sep 17 00:00:00 2001
From: baoshiwei <baoshiwei@shlanbao.cn>
Date: 星期二, 15 四月 2025 13:52:04 +0800
Subject: [PATCH] rjuq
---
detect_onnx.py | 220 +++++++++++
shot_onnx.py | 105 +++++
save_img.py | 31 +
speech/formatmp3.py | 5
replaceLabelNumber.py | 38 ++
quchong.py | 21
speech/deepSpeechTest.py | 20 +
pachong.py | 36 +
paherbbaidu.py | 46 ++
onnx-test.py | 50 ++
speech/whisper2.py | 1
speech/whisperTest.py | 10
onnx_predit.py | 364 +++++++++++++++++++
camera_onnx.py | 101 +++++
pc.py | 35 +
15 files changed, 1,050 insertions(+), 33 deletions(-)
diff --git a/camera_onnx.py b/camera_onnx.py
new file mode 100644
index 0000000..ea02d93
--- /dev/null
+++ b/camera_onnx.py
@@ -0,0 +1,101 @@
+import cv2
+import time
+import numpy as np
+import onnxruntime
+from scipy.special import softmax
+
+# 鍔犺浇ONNX妯″瀷
+session = onnxruntime.InferenceSession("model/classify/s.onnx")
+# 鎽勫儚澶寸储寮曞彿锛岄�氬父涓�0琛ㄧず绗竴涓憚鍍忓ご
+camera_index = 0
+
+# 鎵撳紑鎽勫儚澶�
+cap = cv2.VideoCapture(camera_index, cv2.CAP_DSHOW)
+# 璁剧疆鍒嗚鲸鐜�
+cap.set(cv2.CAP_PROP_FRAME_WIDTH, 3840) # 瀹藉害
+cap.set(cv2.CAP_PROP_FRAME_HEIGHT, 2160) # 楂樺害
+# 妫�鏌ユ憚鍍忓ご鏄惁鎴愬姛鎵撳紑
+if not cap.isOpened():
+ print("鏃犳硶鎵撳紑鎽勫儚澶�")
+ exit()
+
+width = cap.get(cv2.CAP_PROP_FRAME_WIDTH)
+height = cap.get(cv2.CAP_PROP_FRAME_HEIGHT)
+print("鎽勫儚澶村垎杈ㄧ巼:", width, "x", height)
+
+
+
+# 浠巖es.json涓鍙栫被鍒�
+with open("res1-2.json", "r") as f:
+ classes = eval(f.read())
+
+
+
+# 鐩爣鍥惧儚灏哄
+target_width = 1024
+target_height = 768
+
+# 璁℃椂鍣�
+start_time = time.time()
+
+# 寰幆璇诲彇鎽勫儚澶寸敾闈�
+while True:
+ ret, frame = cap.read()
+
+ if not ret:
+ print("鏃犳硶璇诲彇鎽勫儚澶寸敾闈�")
+ break
+
+ # 1920*1080鐨勫浘鍍忥紝涓績瑁佸壀640*480鐨勫尯鍩�
+ cropped_frame = frame[int(height / 2 - target_height / 2):int(height / 2 + target_height / 2),
+ int(width / 2 - target_width / 2):int(width / 2 + target_width / 2)]
+ # 璋冩暣鍥惧儚灏哄
+ resized_frame = cv2.resize(cropped_frame, (target_width, target_height))
+
+ # 鑾峰彇褰撳墠鏃堕棿
+ current_time = time.time()
+
+ #濡傛灉璺濈涓婁竴娆′繚瀛樺凡缁忚繃鍘�1绉掞紝鍒欎繚瀛樺綋鍓嶇敾闈�
+ # if current_time - start_time >= 3.0:
+ # # 鐢熸垚淇濆瓨鏂囦欢鍚嶏紝浠ュ綋鍓嶆椂闂村懡鍚�
+ # save_name = time.strftime("%Y%m%d%H%M%S", time.localtime()) + ".jpg"
+ # # 淇濆瓨璋冩暣灏哄鍚庣殑鍥剧墖
+ # cv2.imwrite(save_path + save_name, frame)
+ # print("淇濆瓨鍥剧墖:", save_name)
+ # # 閲嶇疆璁℃椂鍣�
+ # start_time = time.time()
+
+ # 棰勫鐞�
+ blob = cv2.dnn.blobFromImage(resized_frame, 1 / 255.0, (640, 640), swapRB=True, crop=False)
+
+ # 妯″瀷鎺ㄧ悊
+ outputs = session.run(None, {session.get_inputs()[0].name: blob})
+
+ # print(outputs)
+ # 搴旂敤softmax鍑芥暟
+ probabilities = outputs[0]
+
+ # 鎵惧埌鏈�澶ф鐜囩殑绫诲埆
+ predicted_class = np.argmax(probabilities, axis=1)[0]
+ max_probability = np.max(probabilities, axis=1)[0]
+
+ # 鎵惧埌姒傜巼杈冮珮鐨勫墠鍗佷釜绫诲埆
+ top_ten_classes = np.argsort(probabilities, axis=1)[0][-5:]
+
+ # 杈撳嚭鍓嶅崄涓被鍒�
+ print("Top 5 Classes:")
+ for i in top_ten_classes:
+ print(f"{classes[i]}: {probabilities[0][i]}")
+
+ # 鏄剧ず鐢婚潰
+ cv2.imshow("Camera", resized_frame)
+
+ # 妫�娴嬫寜閿紝濡傛灉鎸変笅q閿垯閫�鍑哄惊鐜�
+ if cv2.waitKey(1) & 0xFF == ord('q'):
+ break
+
+# 鍏抽棴鎽勫儚澶�
+cap.release()
+
+# 鍏抽棴鎵�鏈夌獥鍙�
+cv2.destroyAllWindows()
diff --git a/detect_onnx.py b/detect_onnx.py
new file mode 100644
index 0000000..8d15e9d
--- /dev/null
+++ b/detect_onnx.py
@@ -0,0 +1,220 @@
+import time
+import cv2
+import numpy as np
+import onnxruntime
+
+
+class YOLOv8:
+
+ def __init__(self, path, conf_thres=0.7, iou_thres=0.7):
+ self.conf_threshold = conf_thres
+ self.iou_threshold = iou_thres
+
+ # Initialize model
+ self.initialize_model(path)
+
+ def __call__(self, image):
+ return self.detect_objects(image)
+
+ def initialize_model(self, path):
+ self.session = onnxruntime.InferenceSession(path, providers=['CUDAExecutionProvider', 'CPUExecutionProvider'])
+ # Get model info
+ self.get_input_details()
+ self.get_output_details()
+
+ def detect_objects(self, image):
+ input_tensor, ratio = self.prepare_input(image)
+
+ # Perform inference on the image
+ outputs = self.inference(input_tensor)
+
+ self.boxes, self.scores, self.class_ids = self.process_output(outputs, ratio)
+
+ return self.boxes, self.scores, self.class_ids
+
+ def prepare_input(self, image):
+ self.img_height, self.img_width = image.shape[:2]
+
+ input_img = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
+
+ # Resize鍥剧墖涓嶈鐩存帴浣跨敤resize锛岄渶瑕佹寜姣斾緥缂╂斁锛岀┖鐧藉尯鍩熷～绌虹函鑹插嵆鍙�
+ input_img, ratio = self.ratioresize(input_img)
+
+ # Scale input pixel values to 0 to 1
+ input_img = input_img / 255.0
+ input_img = input_img.transpose(2, 0, 1)
+ input_tensor = input_img[np.newaxis, :, :, :].astype(np.float32)
+
+ return input_tensor, ratio
+
+ def inference(self, input_tensor):
+ start = time.perf_counter()
+ outputs = self.session.run(self.output_names, {self.input_names[0]: input_tensor})
+
+ # print(f"Inference time: {(time.perf_counter() - start)*1000:.2f} ms")
+ return outputs
+
+ def process_output(self, output, ratio):
+ predictions = np.squeeze(output[0]).T
+
+ # Filter out object confidence scores below threshold
+ scores = np.max(predictions[:, 4:], axis=1)
+ predictions = predictions[scores > self.conf_threshold, :]
+ scores = scores[scores > self.conf_threshold]
+
+ if len(scores) == 0:
+ return [], [], []
+
+ # Get the class with the highest confidence
+ class_ids = np.argmax(predictions[:, 4:], axis=1)
+
+ # Get bounding boxes for each object
+ boxes = self.extract_boxes(predictions, ratio)
+
+ # Apply non-maxima suppression to suppress weak, overlapping bounding boxes
+ indices = self.nms(boxes, scores, self.iou_threshold)
+
+ return boxes[indices], scores[indices], class_ids[indices]
+
+ def extract_boxes(self, predictions, ratio):
+ # Extract boxes from predictions
+ boxes = predictions[:, :4]
+
+ # Scale boxes to original image dimensions
+ # boxes = self.rescale_boxes(boxes)
+ boxes *= ratio
+
+ # Convert boxes to xyxy format
+ boxes = self.xywh2xyxy(boxes)
+
+ return boxes
+
+ def rescale_boxes(self, boxes):
+
+ # Rescale boxes to original image dimensions
+
+ input_shape = np.array([self.input_width, self.input_height, self.input_width, self.input_height])
+ boxes = np.divide(boxes, input_shape, dtype=np.float32)
+ boxes *= np.array([self.img_width, self.img_height, self.img_width, self.img_height])
+
+ return boxes
+
+ def get_input_details(self):
+ model_inputs = self.session.get_inputs()
+ self.input_names = [model_inputs[i].name for i in range(len(model_inputs))]
+
+ self.input_shape = model_inputs[0].shape
+ self.input_height = self.input_shape[2]
+ self.input_width = self.input_shape[3]
+
+ def get_output_details(self):
+ model_outputs = self.session.get_outputs()
+ self.output_names = [model_outputs[i].name for i in range(len(model_outputs))]
+
+ # 绛夋瘮渚嬬缉鏀惧浘鐗�
+ def ratioresize(self, im, color=114):
+ shape = im.shape[:2]
+ new_h, new_w = self.input_height, self.input_width
+ padded_img = np.ones((new_h, new_w, 3), dtype=np.uint8) * color
+
+ # Scale ratio (new / old)
+ r = min(new_h / shape[0], new_w / shape[1])
+
+ # Compute padding
+ new_unpad = int(round(shape[1] * r)), int(round(shape[0] * r))
+
+ if shape[::-1] != new_unpad:
+ im = cv2.resize(im, new_unpad, interpolation=cv2.INTER_LINEAR)
+
+ padded_img[: new_unpad[1], : new_unpad[0]] = im
+ padded_img = np.ascontiguousarray(padded_img)
+ return padded_img, 1 / r
+
+ def nms(self, boxes, scores, iou_threshold):
+ # Sort by score
+ sorted_indices = np.argsort(scores)[::-1]
+
+ keep_boxes = []
+ while sorted_indices.size > 0:
+ # Pick the last box
+ box_id = sorted_indices[0]
+ keep_boxes.append(box_id)
+
+ # Compute IoU of the picked box with the rest
+ ious = self.compute_iou(boxes[box_id, :], boxes[sorted_indices[1:], :])
+
+ # Remove boxes with IoU over the threshold
+ keep_indices = np.where(ious < iou_threshold)[0]
+
+ # print(keep_indices.shape, sorted_indices.shape)
+ sorted_indices = sorted_indices[keep_indices + 1]
+
+ return keep_boxes
+
+ def compute_iou(self, box, boxes):
+ # Compute xmin, ymin, xmax, ymax for both boxes
+ xmin = np.maximum(box[0], boxes[:, 0])
+ ymin = np.maximum(box[1], boxes[:, 1])
+ xmax = np.minimum(box[2], boxes[:, 2])
+ ymax = np.minimum(box[3], boxes[:, 3])
+
+ # Compute intersection area
+ intersection_area = np.maximum(0, xmax - xmin) * np.maximum(0, ymax - ymin)
+
+ # Compute union area
+ box_area = (box[2] - box[0]) * (box[3] - box[1])
+ boxes_area = (boxes[:, 2] - boxes[:, 0]) * (boxes[:, 3] - boxes[:, 1])
+ union_area = box_area + boxes_area - intersection_area
+
+ # Compute IoU
+ iou = intersection_area / union_area
+
+ return iou
+
+ def xywh2xyxy(self, x):
+ # Convert bounding box (x, y, w, h) to bounding box (x1, y1, x2, y2)
+ y = np.copy(x)
+ y[..., 0] = x[..., 0] - x[..., 2] / 2
+ y[..., 1] = x[..., 1] - x[..., 3] / 2
+ y[..., 2] = x[..., 0] + x[..., 2] / 2
+ y[..., 3] = x[..., 1] + x[..., 3] / 2
+ return y
+
+
+if __name__ == "__main__":
+ yolov8_detector = YOLOv8('model/detect/best.onnx', conf_thres=0.7, iou_thres=0.7)
+
+
+ # 鎽勫儚澶寸储寮曞彿锛岄�氬父涓�0琛ㄧず绗竴涓憚鍍忓ご
+ camera_index = 0
+
+ # 鎵撳紑鎽勫儚澶�
+ cap = cv2.VideoCapture(camera_index, cv2.CAP_DSHOW)
+ # 璁剧疆鍒嗚鲸鐜�
+ cap.set(cv2.CAP_PROP_FRAME_WIDTH, 3840) # 瀹藉害
+ cap.set(cv2.CAP_PROP_FRAME_HEIGHT, 2160) # 楂樺害
+ # 妫�鏌ユ憚鍍忓ご鏄惁鎴愬姛鎵撳紑
+ if not cap.isOpened():
+ print("鏃犳硶鎵撳紑鎽勫儚澶�")
+ exit()
+ width = cap.get(cv2.CAP_PROP_FRAME_WIDTH)
+ height = cap.get(cv2.CAP_PROP_FRAME_HEIGHT)
+ print("鎽勫儚澶村垎杈ㄧ巼:", width, "x", height)
+ # 鐩爣鍥惧儚灏哄
+ target_width = 1024
+ target_height = 768
+ # 寰幆璇诲彇鎽勫儚澶寸敾闈�
+ while True:
+ ret, frame = cap.read()
+
+ if not ret:
+ print("鏃犳硶璇诲彇鎽勫儚澶寸敾闈�")
+ break
+
+ # 1920*1080鐨勫浘鍍忥紝涓績瑁佸壀640*480鐨勫尯鍩�
+ cropped_frame = frame[int(height / 2 - target_height / 2):int(height / 2 + target_height / 2),
+ int(width / 2 - target_width / 2):int(width / 2 + target_width / 2)]
+ # 璋冩暣鍥惧儚灏哄
+ resized_frame = cv2.resize(cropped_frame, (target_width, target_height))
+ boxes, scores, class_ids = yolov8_detector(resized_frame)
+ print(boxes, scores, class_ids)
\ No newline at end of file
diff --git a/onnx-test.py b/onnx-test.py
new file mode 100644
index 0000000..a654632
--- /dev/null
+++ b/onnx-test.py
@@ -0,0 +1,50 @@
+import cv2
+import numpy as np
+import onnxruntime
+from scipy.special import softmax
+
+# 鍔犺浇ONNX妯″瀷
+session = onnxruntime.InferenceSession("model/classify/best.onnx")
+
+# 璇诲彇鍥剧墖
+img = cv2.imread("D:\\temp\\15.jpg")
+
+# 浠巖es.json涓鍙栫被鍒�
+with open("res1-2.json", "r") as f:
+ classes = eval(f.read())
+
+# 棰勫鐞�
+blob = cv2.dnn.blobFromImage(img, 1/255.0, (640, 640), swapRB=True, crop=False)
+
+# 妯″瀷鎺ㄧ悊
+outputs = session.run(None, {session.get_inputs()[0].name: blob})
+
+# print(outputs)
+# 搴旂敤softmax鍑芥暟
+probabilities = outputs[0]
+
+# 鎵惧埌鏈�澶ф鐜囩殑绫诲埆
+predicted_class = np.argmax(probabilities, axis=1)[0]
+max_probability = np.max(probabilities, axis=1)[0]
+
+# 鎵惧埌姒傜巼杈冮珮鐨勫墠鍗佷釜绫诲埆
+top_ten_classes = np.argsort(probabilities, axis=1)[0][-5:]
+
+# 杈撳嚭鍓嶅崄涓被鍒�
+print("Top 10 Classes:")
+for i in top_ten_classes:
+ print(f"{classes[i]}: {probabilities[0][i]}")
+
+
+
+# 鍚庡鐞�
+# for detection in outputs[0][0]:
+# confidence = detection[4]
+# if confidence > 0.5:
+# class_id = int(detection[5])
+# x, y, w, h = detection[:4]
+# cv2.rectangle(img, (int(x), int(y)), (int(x+w), int(y+h)), (0, 255, 0), 2)
+#
+# # 鏄剧ず缁撴灉
+# cv2.imshow("YOLOv8 Detection", img)
+# cv2.waitKey(0)
\ No newline at end of file
diff --git a/onnx_predit.py b/onnx_predit.py
new file mode 100644
index 0000000..3168a19
--- /dev/null
+++ b/onnx_predit.py
@@ -0,0 +1,364 @@
+import time
+
+import cv2
+import onnxruntime as ort
+from PIL import Image
+import numpy as np
+
+# 缃俊搴�
+confidence_thres = 0.35
+# iou闃堝��
+iou_thres = 0.5
+# 绫诲埆
+
+classes = {0: 'herb'}
+# 闅忔満棰滆壊
+color_palette = np.random.uniform(100, 255, size=(len(classes), 3))
+
+# 鍒ゆ柇鏄娇鐢℅PU鎴朇PU
+providers = [
+ ('CUDAExecutionProvider', {
+ 'device_id': 0, # 鍙互閫夋嫨GPU璁惧ID锛屽鏋滀綘鏈夊涓狦PU
+ }),
+ 'CPUExecutionProvider', # 涔熷彲浠ヨ缃瓹PU浣滀负澶囬��
+]
+
+def calculate_iou(box, other_boxes):
+ """
+ 璁＄畻缁欏畾杈圭晫妗嗕笌涓�缁勫叾浠栬竟鐣屾涔嬮棿鐨勪氦骞舵瘮锛圛oU锛夈��
+
+ 鍙傛暟锛�
+ - box: 鍗曚釜杈圭晫妗嗭紝鏍煎紡涓� [x1, y1, width, height]銆�
+ - other_boxes: 鍏朵粬杈圭晫妗嗙殑鏁扮粍锛屾瘡涓竟鐣屾鐨勬牸寮忎篃涓� [x1, y1, width, height]銆�
+
+ 杩斿洖鍊硷細
+ - iou: 涓�涓暟缁勶紝鍖呭惈缁欏畾杈圭晫妗嗕笌姣忎釜鍏朵粬杈圭晫妗嗙殑IoU鍊笺��
+ """
+
+ # 璁＄畻浜ら泦鐨勫乏涓婅鍧愭爣
+ x1 = np.maximum(box[0], np.array(other_boxes)[:, 0])
+ y1 = np.maximum(box[1], np.array(other_boxes)[:, 1])
+ # 璁＄畻浜ら泦鐨勫彸涓嬭鍧愭爣
+ x2 = np.minimum(box[0] + box[2], np.array(other_boxes)[:, 0] + np.array(other_boxes)[:, 2])
+ y2 = np.minimum(box[1] + box[3], np.array(other_boxes)[:, 1] + np.array(other_boxes)[:, 3])
+ # 璁＄畻浜ら泦鍖哄煙鐨勯潰绉�
+ intersection_area = np.maximum(0, x2 - x1) * np.maximum(0, y2 - y1)
+ # 璁＄畻缁欏畾杈圭晫妗嗙殑闈㈢Н
+ box_area = box[2] * box[3]
+ # 璁＄畻鍏朵粬杈圭晫妗嗙殑闈㈢Н
+ other_boxes_area = np.array(other_boxes)[:, 2] * np.array(other_boxes)[:, 3]
+ # 璁＄畻IoU鍊�
+ iou = intersection_area / (box_area + other_boxes_area - intersection_area)
+ return iou
+
+def custom_NMSBoxes(boxes, scores, confidence_threshold, iou_threshold):
+ # 濡傛灉娌℃湁杈圭晫妗嗭紝鍒欑洿鎺ヨ繑鍥炵┖鍒楄〃
+ if len(boxes) == 0:
+ return []
+ # 灏嗗緱鍒嗗拰杈圭晫妗嗚浆鎹负NumPy鏁扮粍
+ scores = np.array(scores)
+ boxes = np.array(boxes)
+ # 鏍规嵁缃俊搴﹂槇鍊艰繃婊よ竟鐣屾
+ mask = scores > confidence_threshold
+ filtered_boxes = boxes[mask]
+ filtered_scores = scores[mask]
+ # 濡傛灉杩囨护鍚庢病鏈夎竟鐣屾锛屽垯杩斿洖绌哄垪琛�
+ if len(filtered_boxes) == 0:
+ return []
+ # 鏍规嵁缃俊搴﹀緱鍒嗗杈圭晫妗嗚繘琛屾帓搴�
+ sorted_indices = np.argsort(filtered_scores)[::-1]
+ # 鍒濆鍖栦竴涓┖鍒楄〃鏉ュ瓨鍌ㄩ�夋嫨鐨勮竟鐣屾绱㈠紩
+ indices = []
+ # 褰撹繕鏈夋湭澶勭悊鐨勮竟鐣屾鏃讹紝寰幆缁х画
+ while len(sorted_indices) > 0:
+ # 閫夋嫨寰楀垎鏈�楂樼殑杈圭晫妗嗙储寮�
+ current_index = sorted_indices[0]
+ indices.append(current_index)
+ # 濡傛灉鍙墿涓�涓竟鐣屾锛屽垯缁撴潫寰幆
+ if len(sorted_indices) == 1:
+ break
+ # 鑾峰彇褰撳墠杈圭晫妗嗗拰鍏朵粬杈圭晫妗�
+ current_box = filtered_boxes[current_index]
+ other_boxes = filtered_boxes[sorted_indices[1:]]
+ # 璁＄畻褰撳墠杈圭晫妗嗕笌鍏朵粬杈圭晫妗嗙殑IoU
+ iou = calculate_iou(current_box, other_boxes)
+ # 鎵惧埌IoU浣庝簬闃堝�肩殑杈圭晫妗嗭紝鍗充笌褰撳墠杈圭晫妗嗕笉閲嶅彔鐨勮竟鐣屾
+ non_overlapping_indices = np.where(iou <= iou_threshold)[0]
+ # 鏇存柊sorted_indices浠ヤ粎鍖呭惈涓嶉噸鍙犵殑杈圭晫妗�
+ sorted_indices = sorted_indices[non_overlapping_indices + 1]
+ # 杩斿洖閫夋嫨鐨勮竟鐣屾绱㈠紩
+ return indices
+
+
+def draw_detections(img, box, score, class_id):
+ """
+ 鍦ㄨ緭鍏ュ浘鍍忎笂缁樺埗妫�娴嬪埌鐨勫璞＄殑杈圭晫妗嗗拰鏍囩銆�
+
+ 鍙傛暟:
+ img: 瑕佸湪鍏朵笂缁樺埗妫�娴嬬粨鏋滅殑杈撳叆鍥惧儚銆�
+ box: 妫�娴嬪埌鐨勮竟鐣屾銆�
+ score: 瀵瑰簲鐨勬娴嬪緱鍒嗐��
+ class_id: 妫�娴嬪埌鐨勫璞＄殑绫诲埆ID銆�
+
+ 杩斿洖:
+ 鏃�
+ """
+
+ # 鎻愬彇杈圭晫妗嗙殑鍧愭爣
+ x1, y1, w, h = box
+ # 鏍规嵁绫诲埆ID妫�绱㈤鑹�
+ color = color_palette[class_id]
+ # 鍦ㄥ浘鍍忎笂缁樺埗杈圭晫妗�
+ cv2.rectangle(img, (int(x1), int(y1)), (int(x1 + w), int(y1 + h)), color, 2)
+ # 鍒涘缓鏍囩鏂囨湰锛屽寘鎷被鍚嶅拰寰楀垎
+ label = f'{classes[class_id]}: {score:.2f}'
+ # 璁＄畻鏍囩鏂囨湰鐨勫昂瀵�
+ (label_width, label_height), _ = cv2.getTextSize(label, cv2.FONT_HERSHEY_SIMPLEX, 0.5, 1)
+ # 璁＄畻鏍囩鏂囨湰鐨勪綅缃�
+ label_x = x1
+ label_y = y1 - 10 if y1 - 10 > label_height else y1 + 10
+ # 缁樺埗濉厖鐨勭煩褰綔涓烘爣绛炬枃鏈殑鑳屾櫙
+ cv2.rectangle(img, (label_x, label_y - label_height), (label_x + label_width, label_y + label_height), color, cv2.FILLED)
+ # 鍦ㄥ浘鍍忎笂缁樺埗鏍囩鏂囨湰
+ cv2.putText(img, label, (label_x, label_y), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 0, 0), 1, cv2.LINE_AA)
+
+
+def preprocess(img, input_width, input_height):
+ """
+ 鍦ㄦ墽琛屾帹鐞嗕箣鍓嶉澶勭悊杈撳叆鍥惧儚銆�
+
+ 杩斿洖:
+ image_data: 涓烘帹鐞嗗噯澶囧ソ鐨勯澶勭悊鍚庣殑鍥惧儚鏁版嵁銆�
+ """
+
+ # 鑾峰彇杈撳叆鍥惧儚鐨勯珮搴﹀拰瀹藉害
+ img_height, img_width = img.shape[:2]
+ # 灏嗗浘鍍忛鑹茬┖闂翠粠BGR杞崲涓篟GB
+ img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
+ # 灏嗗浘鍍忓ぇ灏忚皟鏁翠负鍖归厤杈撳叆褰㈢姸
+ img = cv2.resize(img, (input_width, input_height))
+ # 閫氳繃闄や互255.0鏉ュ綊涓�鍖栧浘鍍忔暟鎹�
+ image_data = np.array(img) / 255.0
+ # 杞疆鍥惧儚锛屼娇閫氶亾缁村害涓虹涓�缁�
+ image_data = np.transpose(image_data, (2, 0, 1)) # 閫氶亾棣�
+ # 鎵╁睍鍥惧儚鏁版嵁鐨勭淮搴︿互鍖归厤棰勬湡鐨勮緭鍏ュ舰鐘�
+ image_data = np.expand_dims(image_data, axis=0).astype(np.float32)
+ # 杩斿洖棰勫鐞嗗悗鐨勫浘鍍忔暟鎹�
+ return image_data, img_height, img_width
+
+def postprocess(input_image, output, input_width, input_height, img_width, img_height):
+ """
+ 瀵规ā鍨嬭緭鍑鸿繘琛屽悗澶勭悊锛屾彁鍙栬竟鐣屾銆佸緱鍒嗗拰绫诲埆ID銆�
+
+ 鍙傛暟:
+ input_image (numpy.ndarray): 杈撳叆鍥惧儚銆�
+ output (numpy.ndarray): 妯″瀷鐨勮緭鍑恒��
+ input_width (int): 妯″瀷杈撳叆瀹藉害銆�
+ input_height (int): 妯″瀷杈撳叆楂樺害銆�
+ img_width (int): 鍘熷鍥惧儚瀹藉害銆�
+ img_height (int): 鍘熷鍥惧儚楂樺害銆�
+
+ 杩斿洖:
+ numpy.ndarray: 缁樺埗浜嗘娴嬬粨鏋滅殑杈撳叆鍥惧儚銆�
+ """
+
+ # 杞疆鍜屽帇缂╄緭鍑轰互鍖归厤棰勬湡鐨勫舰鐘�
+ outputs = np.transpose(np.squeeze(output[0]))
+ # 鑾峰彇杈撳嚭鏁扮粍鐨勮鏁�
+ rows = outputs.shape[0]
+ # 鐢ㄤ簬瀛樺偍妫�娴嬬殑杈圭晫妗嗐�佸緱鍒嗗拰绫诲埆ID鐨勫垪琛�
+ boxes = []
+ scores = []
+ class_ids = []
+ # 璁＄畻杈圭晫妗嗗潗鏍囩殑缂╂斁鍥犲瓙
+ x_factor = img_width / input_width
+ y_factor = img_height / input_height
+ # 閬嶅巻杈撳嚭鏁扮粍鐨勬瘡涓�琛�
+ for i in range(rows):
+ # 浠庡綋鍓嶈鎻愬彇绫诲埆寰楀垎
+ classes_scores = outputs[i][4:]
+ # 鎵惧埌绫诲埆寰楀垎涓殑鏈�澶у緱鍒�
+ max_score = np.amax(classes_scores)
+ # 濡傛灉鏈�澶у緱鍒嗛珮浜庣疆淇″害闃堝��
+ if max_score >= confidence_thres:
+ # 鑾峰彇寰楀垎鏈�楂樼殑绫诲埆ID
+ class_id = np.argmax(classes_scores)
+ # 浠庡綋鍓嶈鎻愬彇杈圭晫妗嗗潗鏍�
+ x, y, w, h = outputs[i][0], outputs[i][1], outputs[i][2], outputs[i][3]
+ # 璁＄畻杈圭晫妗嗙殑缂╂斁鍧愭爣
+ left = int((x - w / 2) * x_factor)
+ top = int((y - h / 2) * y_factor)
+ width = int(w * x_factor)
+ height = int(h * y_factor)
+ # 灏嗙被鍒獻D銆佸緱鍒嗗拰妗嗗潗鏍囨坊鍔犲埌鍚勮嚜鐨勫垪琛ㄤ腑
+ class_ids.append(class_id)
+ scores.append(max_score)
+ boxes.append([left, top, width, height])
+ # 搴旂敤闈炴渶澶ф姂鍒惰繃婊ら噸鍙犵殑杈圭晫妗�
+ indices = custom_NMSBoxes(boxes, scores, confidence_thres, iou_thres)
+ # 閬嶅巻闈炴渶澶ф姂鍒跺悗鐨勯�夊畾绱㈠紩
+ for i in indices:
+ # 鏍规嵁绱㈠紩鑾峰彇妗嗐�佸緱鍒嗗拰绫诲埆ID
+ box = boxes[i]
+ score = scores[i]
+ class_id = class_ids[i]
+ # 鍦ㄨ緭鍏ュ浘鍍忎笂缁樺埗妫�娴嬬粨鏋�
+ draw_detections(input_image, box, score, class_id)
+ # 杩斿洖淇敼鍚庣殑杈撳叆鍥惧儚
+ return input_image
+
+def init_detect_model(model_path):
+ # 浣跨敤ONNX妯″瀷鏂囦欢鍒涘缓涓�涓帹鐞嗕細璇濓紝骞舵寚瀹氭墽琛屾彁渚涜��
+ session = ort.InferenceSession(model_path, providers=providers)
+ # 鑾峰彇妯″瀷鐨勮緭鍏ヤ俊鎭�
+ model_inputs = session.get_inputs()
+ # 鑾峰彇杈撳叆鐨勫舰鐘讹紝鐢ㄤ簬鍚庣画浣跨敤
+ input_shape = model_inputs[0].shape
+ # 浠庤緭鍏ュ舰鐘朵腑鎻愬彇杈撳叆瀹藉害
+ input_width = input_shape[2]
+ # 浠庤緭鍏ュ舰鐘朵腑鎻愬彇杈撳叆楂樺害
+ input_height = input_shape[3]
+ # 杩斿洖浼氳瘽銆佹ā鍨嬭緭鍏ヤ俊鎭�佽緭鍏ュ搴﹀拰杈撳叆楂樺害
+ return session, model_inputs, input_width, input_height
+
+def detect_object(image, session, model_inputs, input_width, input_height):
+ # 濡傛灉杈撳叆鐨勫浘鍍忔槸PIL鍥惧儚瀵硅薄锛屽皢鍏惰浆鎹负NumPy鏁扮粍
+ if isinstance(image, Image.Image):
+ result_image = np.array(image)
+ else:
+ # 鍚﹀垯锛岀洿鎺ヤ娇鐢ㄨ緭鍏ョ殑鍥惧儚锛堝亣瀹氬凡缁忔槸NumPy鏁扮粍锛�
+ result_image = image
+ # 棰勫鐞嗗浘鍍忔暟鎹紝璋冩暣鍥惧儚澶у皬骞跺彲鑳借繘琛屽綊涓�鍖栫瓑鎿嶄綔
+ img_data, img_height, img_width = preprocess(result_image, input_width, input_height)
+ # 浣跨敤棰勫鐞嗗悗鐨勫浘鍍忔暟鎹繘琛屾帹鐞�
+ outputs = session.run(None, {model_inputs[0].name: img_data})
+ # 瀵规帹鐞嗙粨鏋滆繘琛屽悗澶勭悊锛屼緥濡傝В鐮佹娴嬫锛岃繃婊や綆缃俊搴︾殑妫�娴嬬瓑
+ output_image = postprocess(result_image, outputs, input_width, input_height, img_width, img_height)
+ # 杩斿洖澶勭悊鍚庣殑鍥惧儚
+ return output_image
+if __name__ == '__main__':
+ # 妯″瀷鏂囦欢鐨勮矾寰�
+ model_path = 'model/detect/best.onnx'
+ # 鍒濆鍖栨娴嬫ā鍨嬶紝鍔犺浇妯″瀷骞惰幏鍙栨ā鍨嬭緭鍏ヨ妭鐐逛俊鎭拰杈撳叆鍥惧儚鐨勫搴︺�侀珮搴�
+ session, model_inputs, input_width, input_height = init_detect_model(model_path)
+ # 涓夌妯″紡 1涓哄浘鐗囬娴嬶紝骞舵樉绀虹粨鏋滃浘鐗囷紱2涓烘憚鍍忓ご妫�娴嬶紝骞跺疄鏃舵樉绀篎PS锛� 3涓鸿棰戞娴嬶紝骞朵繚瀛樼粨鏋滆棰�
+ mode = 2
+ if mode == 1:
+ # 璇诲彇鍥惧儚鏂囦欢
+ image_data = cv2.imread("street.jpg")
+ # 浣跨敤妫�娴嬫ā鍨嬪璇诲叆鐨勫浘鍍忚繘琛屽璞℃娴�
+ result_image = detect_object(image_data, session, model_inputs, input_width, input_height)
+ # 灏嗘娴嬪悗鐨勫浘鍍忎繚瀛樺埌鏂囦欢
+ cv2.imwrite("output_image.jpg", result_image)
+ # 鍦ㄧ獥鍙ｄ腑鏄剧ず妫�娴嬪悗鐨勫浘鍍�
+ cv2.imshow('Output', result_image)
+ # 绛夊緟鐢ㄦ埛鎸夐敭锛岀劧鍚庡叧闂樉绀虹獥鍙�
+ cv2.waitKey(0)
+ elif mode == 2:
+ # 鎽勫儚澶寸储寮曞彿锛岄�氬父涓�0琛ㄧず绗竴涓憚鍍忓ご
+ camera_index = 0
+
+ # 鎵撳紑鎽勫儚澶�
+ cap = cv2.VideoCapture(camera_index, cv2.CAP_DSHOW)
+ # 璁剧疆鍒嗚鲸鐜�
+ cap.set(cv2.CAP_PROP_FRAME_WIDTH, 3840) # 瀹藉害
+ cap.set(cv2.CAP_PROP_FRAME_HEIGHT, 2160) # 楂樺害
+ # 妫�鏌ユ憚鍍忓ご鏄惁鎴愬姛鎵撳紑
+ if not cap.isOpened():
+ print("鏃犳硶鎵撳紑鎽勫儚澶�")
+ exit()
+ width = cap.get(cv2.CAP_PROP_FRAME_WIDTH)
+ height = cap.get(cv2.CAP_PROP_FRAME_HEIGHT)
+ print("鎽勫儚澶村垎杈ㄧ巼:", width, "x", height)
+
+
+ # 鍒濆鍖栧抚鏁拌鏁板櫒鍜岃捣濮嬫椂闂�
+ frame_count = 0
+ start_time = time.time()
+
+ # 鐩爣鍥惧儚灏哄
+ target_width = 1024
+ target_height = 768
+
+
+
+ # 寰幆璇诲彇鎽勫儚澶磋棰戞祦
+ while True:
+ # 璇诲彇涓�甯�
+ ret, frame = cap.read()
+ # 妫�鏌ュ抚鏄惁鎴愬姛璇诲彇
+ if not ret:
+ print("Error: Could not read frame.")
+ break
+ # 1920*1080鐨勫浘鍍忥紝涓績瑁佸壀640*480鐨勫尯鍩�
+ cropped_frame = frame[int(height / 2 - target_height / 2):int(height / 2 + target_height / 2),
+ int(width / 2 - target_width / 2):int(width / 2 + target_width / 2)]
+ # 璋冩暣鍥惧儚灏哄
+ resized_frame = cv2.resize(cropped_frame, (target_width, target_height))
+
+ # 浣跨敤妫�娴嬫ā鍨嬪璇诲叆鐨勫抚杩涜瀵硅薄妫�娴�
+ output_image = detect_object(resized_frame, session, model_inputs, input_width, input_height)
+ # 璁＄畻甯ч�熺巼
+ frame_count += 1
+ end_time = time.time()
+ elapsed_time = end_time - start_time
+ fps = frame_count / elapsed_time
+ print(f"FPS: {fps:.2f}")
+ # 灏咶PS缁樺埗鍦ㄥ浘鍍忎笂
+ cv2.putText(output_image, f"FPS: {fps:.2f}", (10, 30), cv2.FONT_HERSHEY_SIMPLEX, 1, (255, 255, 255), 2, cv2.LINE_AA)
+ # 鍦ㄧ獥鍙ｄ腑鏄剧ず褰撳墠甯�
+ cv2.imshow("Video", output_image)
+ # 鎸変笅 'q' 閿��鍑哄惊鐜�
+ if cv2.waitKey(1) & 0xFF == ord('q'):
+ break
+ # 閲婃斁鎽勫儚澶磋祫婧�
+ cap.release()
+ # 鍏抽棴绐楀彛
+ cv2.destroyAllWindows()
+ elif mode == 3:
+ # 杈撳叆瑙嗛璺緞
+ input_video_path = 'kun.mp4'
+ # 杈撳嚭瑙嗛璺緞
+ output_video_path = 'kun_det.mp4'
+ # 鎵撳紑瑙嗛鏂囦欢
+ cap = cv2.VideoCapture(input_video_path)
+ # 妫�鏌ヨ棰戞槸鍚︽垚鍔熸墦寮�
+ if not cap.isOpened():
+ print("Error: Could not open video.")
+ exit()
+ # 璇诲彇瑙嗛鐨勫熀鏈俊鎭�
+ frame_width = int(cap.get(3))
+ frame_height = int(cap.get(4))
+ fps = cap.get(cv2.CAP_PROP_FPS)
+ # 瀹氫箟瑙嗛缂栫爜鍣ㄥ拰鍒涘缓VideoWriter瀵硅薄
+ fourcc = cv2.VideoWriter_fourcc(*'mp4v') # 鏍规嵁鏂囦欢鍚嶅悗缂�浣跨敤鍚堥�傜殑缂栫爜鍣�
+ out = cv2.VideoWriter(output_video_path, fourcc, fps, (frame_width, frame_height))
+ # 鍒濆鍖栧抚鏁拌鏁板櫒鍜岃捣濮嬫椂闂�
+ frame_count = 0
+ start_time = time.time()
+ while True:
+ ret, frame = cap.read()
+ if not ret:
+ print("Info: End of video file.")
+ break
+ # 瀵硅鍏ョ殑甯ц繘琛屽璞℃娴�
+ output_image = detect_object(frame, session, model_inputs, input_width, input_height)
+ # 璁＄畻骞舵墦鍗板抚閫熺巼
+ frame_count += 1
+ end_time = time.time()
+ elapsed_time = end_time - start_time
+ if elapsed_time > 0:
+ fps = frame_count / elapsed_time
+ print(f"FPS: {fps:.2f}")
+ # 灏嗗鐞嗗悗鐨勫抚鍐欏叆杈撳嚭瑙嗛
+ out.write(output_image)
+ #锛堝彲閫夛級瀹炴椂鏄剧ず澶勭悊鍚庣殑瑙嗛甯�
+ cv2.imshow("Output Video", output_image)
+ if cv2.waitKey(1) & 0xFF == ord('q'):
+ break
+ # 閲婃斁璧勬簮
+ cap.release()
+ out.release()
+ cv2.destroyAllWindows()
+ else:
+ print("杈撳叆閿欒锛岃妫�鏌ode鐨勮祴鍊�")
diff --git a/pachong.py b/pachong.py
index 4841de4..d82db8f 100644
--- a/pachong.py
+++ b/pachong.py
@@ -1,13 +1,20 @@
+import concurrent.futures
import os
import time
import requests
import re
+
def imgdata_set(save_path, word, epoch):
+ if not os.path.exists(save_path):
+ os.makedirs(save_path)
+ else:
+ return 0
q = 0 # 鍋滄鐖彇鍥剧墖鏉′欢
a = 0 # 鍥剧墖鍚嶇О
while(True):
time.sleep(1)
+ print("寮�濮嬬埇鍙栧浘鐗�")
url = "https://image.baidu.com/search/flip?tn=baiduimage&ie=utf-8&word={}&pn={}&ct=&ic=0&lm=-1&width=0&height=0".format(word, q)
# word=闇�瑕佹悳绱㈢殑鍚嶅瓧
headers = {
@@ -16,21 +23,28 @@
response = requests.get(url, headers=headers) # 鍙戦�佽姹傝幏鍙栧搷搴�
html = response.text # 鑾峰彇鍝嶅簲鐨凥TML鍐呭
urls = re.findall('"objURL":"(.*?)"', html) # 浣跨敤姝ｅ垯琛ㄨ揪寮忔彁鍙栧浘鐗嘦RL
- for url in urls:
- try:
- print(a) # 鍥剧墖鐨勫悕瀛�
- response = requests.get(url, headers=headers) # 鍙戦�佽姹傝幏鍙栧浘鐗囧搷搴擡:\yaocai\juhua
- image = response.content # 鑾峰彇鍥剧墖鍐呭
- with open(os.path.join(save_path, "{}.jpg".format(a)), 'wb') as f: # 灏嗗浘鐗囧唴瀹逛繚瀛樺埌鎸囧畾璺緞
- f.write(image)
- a = a + 1
- except Exception as e:
- pass
- continue
+ print(len(urls))
+ # 浣跨敤concurrent.futures瀹炵幇骞跺彂涓嬭浇
+ with concurrent.futures.ThreadPoolExecutor( max_workers=10) as executor:
+ # 鎻愪氦鎵�鏈変笅杞戒换鍔″苟鏀堕泦future瀵硅薄
+ futures = [executor.submit(download_image, index, headers,save_path,url ) for index,url in enumerate(urls)]
q = q + 20
if (q / 20) >= int(epoch):
break
+
+def download_image(a, headers, save_path, url):
+ try:
+ print(a) # 鍥剧墖鐨勫悕瀛�
+ response = requests.get(url, headers=headers, timeout=10) # 鍙戦�佽姹傝幏鍙栧浘鐗囧搷搴�
+ # 濡傛灉娌℃湁涓�鐩村搷搴旀�庝箞澶勭悊
+ image = response.content # 鑾峰彇鍥剧墖鍐呭
+ with open(os.path.join(save_path, "{}.jpg".format(a)), 'wb') as f: # 灏嗗浘鐗囧唴瀹逛繚瀛樺埌鎸囧畾璺緞
+ f.write(image)
+ except Exception as e:
+ pass
+
+
if __name__ == "__main__":
save_path = input('浣犳兂淇濆瓨鐨勮矾寰勶細') # 璇㈤棶鐢ㄦ埛淇濆瓨璺緞
word = input('浣犳兂瑕佷笅杞戒粈涔堝浘鐗囷紵璇疯緭鍏�:') # 璇㈤棶鐢ㄦ埛鎼滅储鍏抽敭璇�
diff --git a/paherbbaidu.py b/paherbbaidu.py
new file mode 100644
index 0000000..493ed57
--- /dev/null
+++ b/paherbbaidu.py
@@ -0,0 +1,46 @@
+import mysql.connector
+from mysql.connector import Error
+from pachong import imgdata_set
+from quchong import quchongmethod
+import shutil
+
+
+# 杩炴帴mysql 鏌ヨ鏁版嵁搴�
+def mysql_connect():
+ try:
+ conn = mysql.connector.connect(host='localhost',
+ database='herb',
+ user='root',
+ password='123456')
+ if conn.is_connected():
+ print('Connected to MySQL database')
+ return conn
+ except Error as e:
+ print(e)
+
+# 鏌ヨdry_herb_info琛�
+def mysql_select(conn):
+ cursor = conn.cursor()
+ sql = "SELECT name, pinyin FROM dry_herb_info where pinyin is not null and pinyin <> ''"
+ cursor.execute(sql)
+ result = cursor.fetchall()
+ return result
+
+
+# main
+if __name__ == '__main__':
+ conn = mysql_connect()
+
+ result = mysql_select(conn)
+
+ for row in result:
+ name = row[0]
+ pinyin = row[1]
+ print(name)
+ print(pinyin)
+ imgdata_set('E:/pachong/'+pinyin, name + '楗墖', 2)
+ quchongmethod('E:/pachong/'+pinyin, 'E:/pachong/2/'+pinyin)
+ #鍒犻櫎涓存椂鏂囦欢澶�
+
+ shutil.rmtree('E:/pachong/2/'+pinyin)
+
diff --git a/pc.py b/pc.py
new file mode 100644
index 0000000..75ac312
--- /dev/null
+++ b/pc.py
@@ -0,0 +1,35 @@
+import os
+import requests
+from bs4 import BeautifulSoup
+from PIL import Image
+from io import BytesIO
+
+
+def fetch_images(keyword, save_path):
+ if not os.path.exists(save_path):
+ os.makedirs(save_path)
+
+ url = f"https://image.baidu.com/search/flip?tn=baiduimage&ie=utf-8&word={keyword}"
+ headers = {
+ 'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/58.0.3029.110 Safari/537.3'
+ }
+
+ response = requests.get(url, headers=headers)
+ soup = BeautifulSoup(response.text, 'html.parser')
+ img_tags = soup.find_all('ObjURL', class_='obj')
+
+ for i, img_tag in enumerate(img_tags):
+ try:
+ img_url = img_tag['data-src']
+ img_data = requests.get(img_url).content
+ img = Image.open(BytesIO(img_data))
+ img.save(os.path.join(save_path, f"{keyword}_{i}.jpg"))
+ print(f"Downloaded {keyword}_{i}.jpg")
+ except Exception as e:
+ print(f"Failed to download image: {e}")
+
+
+if __name__ == "__main__":
+ keyword = "浜哄弬"
+ save_path = "images"
+ fetch_images(keyword, save_path)
\ No newline at end of file
diff --git a/quchong.py b/quchong.py
index 3a27d70..d48414c 100644
--- a/quchong.py
+++ b/quchong.py
@@ -52,12 +52,9 @@
result = "涓ゅ紶鍥剧浉鍚�"
return result
-if __name__ == '__main__':
- load_path = 'E:\yaocai\yinyanghuo' # 瑕佸幓閲嶇殑鏂囦欢澶�
- save_path = 'E:\yaocai\\2\\yinyanghuo' # 绌烘枃浠跺す锛岀敤浜庡瓨鍌ㄦ娴嬪埌鐨勯噸澶嶇殑鐓х墖
+def quchongmethod(load_path, save_path):
os.makedirs(save_path, exist_ok=True)
-
# 鑾峰彇鍥剧墖鍒楄〃 file_map锛屽瓧鍏竰鏂囦欢璺緞filename : 鏂囦欢澶у皬image_size}
file_map = {}
image_size = 0
@@ -70,13 +67,11 @@
# print('the full name of the file is %s' % os.path.join(parent, filename))
image_size = os.path.getsize(os.path.join(parent, filename))
file_map.setdefault(os.path.join(parent, filename), image_size)
-
# 鑾峰彇鐨勫浘鐗囧垪琛ㄦ寜 鏂囦欢澶у皬image_size 鎺掑簭
file_map = sorted(file_map.items(), key=lambda d: d[1], reverse=False)
file_list = []
for filename, image_size in file_map:
file_list.append(filename)
-
# 鍙栧嚭閲嶅鐨勫浘鐗�
file_repeat = []
for currIndex, filename in enumerate(file_list):
@@ -93,16 +88,22 @@
except Exception as e:
pass
- if(result == "涓ゅ紶鍥剧浉鍚�"):
+ if (result == "涓ゅ紶鍥剧浉鍚�"):
file_repeat.append(file_list[currIndex + 1])
print("鐩稿悓鐨勫浘鐗囷細", file_list[currIndex], file_list[currIndex + 1])
else:
print('涓嶅悓鐨勫浘鐗囷細', file_list[currIndex], file_list[currIndex + 1])
currIndex += 1
- if currIndex >= (len(file_list)-2):
+ if currIndex >= (len(file_list) - 2):
break
-
# 灏嗛噸澶嶇殑鍥剧墖绉诲姩鍒版柊鐨勬枃浠跺す锛屽疄鐜板鍘熸枃浠跺す闄嶉噸
for image in file_repeat:
shutil.move(image, save_path)
- print("姝ｅ湪绉婚櫎閲嶅鐓х墖锛�", image)
\ No newline at end of file
+ print("姝ｅ湪绉婚櫎閲嶅鐓х墖锛�", image)
+
+
+if __name__ == '__main__':
+
+ load_path = 'E:\yaocai\\chenpi' # 瑕佸幓閲嶇殑鏂囦欢澶�
+ save_path = 'E:\yaocai\\2\\chenpi' # 绌烘枃浠跺す锛岀敤浜庡瓨鍌ㄦ娴嬪埌鐨勯噸澶嶇殑鐓х墖
+ quchongmethod(load_path,save_path)
\ No newline at end of file
diff --git a/replaceLabelNumber.py b/replaceLabelNumber.py
new file mode 100644
index 0000000..0d9ee10
--- /dev/null
+++ b/replaceLabelNumber.py
@@ -0,0 +1,38 @@
+import os
+
+def process_file(file_path):
+ with open(file_path, 'r') as file:
+ lines = file.readlines()
+
+ with open(file_path, 'w') as file:
+ for line in lines:
+ # 灏濊瘯浠庢瘡涓�琛屼腑鎻愬彇绗竴涓暟瀛�
+ try:
+ number = int(line.split()[0])
+ except ValueError:
+ # 濡傛灉鎻愬彇澶辫触锛岃烦杩囧綋鍓嶈
+ file.write(line)
+ continue
+
+
+ new_number = number + 1
+
+ # 鏇挎崲鍘熷琛屼腑鐨勭涓�涓暟瀛�
+ updated_line = line.replace(str(number), str(new_number), 1)
+
+ # 鍐欏叆鏇存柊鍚庣殑琛�
+ file.write(updated_line)
+
+if __name__ == "__main__":
+ # 鎸囧畾鐩綍
+ target_directory = 'E:\\herb_scan.v1i.yolov8\\valid\\labels'
+
+ # 鑾峰彇鐩綍涓嬬殑鎵�鏈塼xt鏂囦欢
+ txt_files = [file for file in os.listdir(target_directory) if file.endswith('.txt')]
+
+ # 澶勭悊姣忎釜鏂囦欢
+ for txt_file in txt_files:
+ file_path = os.path.join(target_directory, txt_file)
+ process_file(file_path)
+
+ print("澶勭悊瀹屾垚锛�")
diff --git a/save_img.py b/save_img.py
index f588ec0..074e83a 100644
--- a/save_img.py
+++ b/save_img.py
@@ -6,12 +6,17 @@
# 鎵撳紑鎽勫儚澶�
cap = cv2.VideoCapture(camera_index)
-
+# 璁剧疆鍒嗚鲸鐜�
+# cap.set(cv2.CAP_PROP_FRAME_WIDTH, 3840) # 瀹藉害
+# cap.set(cv2.CAP_PROP_FRAME_HEIGHT, 2160) # 楂樺害
# 妫�鏌ユ憚鍍忓ご鏄惁鎴愬姛鎵撳紑
if not cap.isOpened():
print("鏃犳硶鎵撳紑鎽勫儚澶�")
exit()
+width = cap.get(cv2.CAP_PROP_FRAME_WIDTH)
+height = cap.get(cv2.CAP_PROP_FRAME_HEIGHT)
+print("鎽勫儚澶村垎杈ㄧ巼:", width, "x", height)
# 鍥剧墖淇濆瓨璺緞
save_path = "captured_images/"
@@ -36,27 +41,29 @@
print("鏃犳硶璇诲彇鎽勫儚澶寸敾闈�")
break
+# 瑁佸壀鍥惧儚
+# cropped_frame = frame[750:1230, 1650:2290]
# 璋冩暣鍥惧儚灏哄
resized_frame = cv2.resize(frame, (target_width, target_height))
# 鑾峰彇褰撳墠鏃堕棿
current_time = time.time()
- # 濡傛灉璺濈涓婁竴娆′繚瀛樺凡缁忚繃鍘�1绉掞紝鍒欎繚瀛樺綋鍓嶇敾闈�
- if current_time - start_time >= 1.0:
- # 鐢熸垚淇濆瓨鏂囦欢鍚嶏紝浠ュ綋鍓嶆椂闂村懡鍚�
- save_name = time.strftime("%Y%m%d%H%M%S", time.localtime()) + ".jpg"
- # 淇濆瓨璋冩暣灏哄鍚庣殑鍥剧墖
- cv2.imwrite(save_path + save_name, resized_frame)
- print("淇濆瓨鍥剧墖:", save_name)
- # 閲嶇疆璁℃椂鍣�
- start_time = time.time()
+ #濡傛灉璺濈涓婁竴娆′繚瀛樺凡缁忚繃鍘�1绉掞紝鍒欎繚瀛樺綋鍓嶇敾闈�
+ # if current_time - start_time >= 3.0:
+ # # 鐢熸垚淇濆瓨鏂囦欢鍚嶏紝浠ュ綋鍓嶆椂闂村懡鍚�
+ # save_name = time.strftime("%Y%m%d%H%M%S", time.localtime()) + ".jpg"
+ # # 淇濆瓨璋冩暣灏哄鍚庣殑鍥剧墖
+ # cv2.imwrite(save_path + save_name, frame)
+ # print("淇濆瓨鍥剧墖:", save_name)
+ # # 閲嶇疆璁℃椂鍣�
+ # start_time = time.time()
# 鏄剧ず鐢婚潰
- cv2.imshow("Camera", resized_frame)
+ cv2.imshow("Camera", frame)
# 妫�娴嬫寜閿紝濡傛灉鎸変笅q閿垯閫�鍑哄惊鐜�
- if cv2.waitKey(1000) & 0xFF == ord('q'):
+ if cv2.waitKey(1) & 0xFF == ord('q'):
break
# 鍏抽棴鎽勫儚澶�
diff --git a/shot_onnx.py b/shot_onnx.py
new file mode 100644
index 0000000..de87634
--- /dev/null
+++ b/shot_onnx.py
@@ -0,0 +1,105 @@
+import cv2
+import time
+import numpy as np
+import onnxruntime
+from scipy.special import softmax
+
+# 鍔犺浇ONNX妯″瀷
+session = onnxruntime.InferenceSession("model/detect/best.onnx")
+# 鎽勫儚澶寸储寮曞彿锛岄�氬父涓�0琛ㄧず绗竴涓憚鍍忓ご
+camera_index = 0
+
+# 鎵撳紑鎽勫儚澶�
+cap = cv2.VideoCapture(camera_index, cv2.CAP_DSHOW)
+# 璁剧疆鍒嗚鲸鐜�
+cap.set(cv2.CAP_PROP_FRAME_WIDTH, 3840) # 瀹藉害
+cap.set(cv2.CAP_PROP_FRAME_HEIGHT, 2160) # 楂樺害
+# 妫�鏌ユ憚鍍忓ご鏄惁鎴愬姛鎵撳紑
+if not cap.isOpened():
+ print("鏃犳硶鎵撳紑鎽勫儚澶�")
+ exit()
+
+width = cap.get(cv2.CAP_PROP_FRAME_WIDTH)
+height = cap.get(cv2.CAP_PROP_FRAME_HEIGHT)
+print("鎽勫儚澶村垎杈ㄧ巼:", width, "x", height)
+
+
+
+# 浠巖es.json涓鍙栫被鍒�
+# with open("res1-2.json", "r") as f:
+# classes = eval(f.read())
+
+
+
+# 鐩爣鍥惧儚灏哄
+target_width = 1024
+target_height = 768
+
+# 璁℃椂鍣�
+start_time = time.time()
+
+# 寰幆璇诲彇鎽勫儚澶寸敾闈�
+while True:
+ ret, frame = cap.read()
+
+ if not ret:
+ print("鏃犳硶璇诲彇鎽勫儚澶寸敾闈�")
+ break
+
+ # 1920*1080鐨勫浘鍍忥紝涓績瑁佸壀640*480鐨勫尯鍩�
+ cropped_frame = frame[int(height / 2 - target_height / 2):int(height / 2 + target_height / 2),
+ int(width / 2 - target_width / 2):int(width / 2 + target_width / 2)]
+ # 璋冩暣鍥惧儚灏哄
+ resized_frame = cv2.resize(cropped_frame, (target_width, target_height))
+
+ # 鑾峰彇褰撳墠鏃堕棿
+ current_time = time.time()
+
+ #濡傛灉璺濈涓婁竴娆′繚瀛樺凡缁忚繃鍘�1绉掞紝鍒欎繚瀛樺綋鍓嶇敾闈�
+ # if current_time - start_time >= 3.0:
+ # # 鐢熸垚淇濆瓨鏂囦欢鍚嶏紝浠ュ綋鍓嶆椂闂村懡鍚�
+ # save_name = time.strftime("%Y%m%d%H%M%S", time.localtime()) + ".jpg"
+ # # 淇濆瓨璋冩暣灏哄鍚庣殑鍥剧墖
+ # cv2.imwrite(save_path + save_name, frame)
+ # print("淇濆瓨鍥剧墖:", save_name)
+ # # 閲嶇疆璁℃椂鍣�
+ # start_time = time.time()
+
+ # 棰勫鐞�
+ blob = cv2.dnn.blobFromImage(resized_frame, 1 / 255.0, (640, 640), swapRB=True, crop=False)
+
+ # 妯″瀷鎺ㄧ悊
+ outputs = session.run(None, {session.get_inputs()[0].name: blob})
+
+ output = np.transpose(np.squeeze(outputs[0]));
+ rows = output.shape[0]
+ boxes = []
+ scores = []
+ class_ids = []
+ for i in range(rows):
+ classes_scores = output[i][4:]
+ max_score = np.amax(classes_scores)
+ if max_score > 0.5:
+ classid = np.argmax(classes_scores)
+ scores.append(max_score)
+ class_ids.append(classid)
+
+ print(class_ids)
+ print(scores)
+ # # 杈撳嚭鍓嶅崄涓被鍒�
+ # print("Top 5 Classes:")
+ # for i in top_ten_classes:
+ # print(f"{classes[i]}: {probabilities[0][i]}")
+
+ # 鏄剧ず鐢婚潰
+ cv2.imshow("Camera", resized_frame)
+
+ # 妫�娴嬫寜閿紝濡傛灉鎸変笅q閿垯閫�鍑哄惊鐜�
+ if cv2.waitKey(1) & 0xFF == ord('q'):
+ break
+
+# 鍏抽棴鎽勫儚澶�
+cap.release()
+
+# 鍏抽棴鎵�鏈夌獥鍙�
+cv2.destroyAllWindows()
diff --git a/speech/deepSpeechTest.py b/speech/deepSpeechTest.py
new file mode 100644
index 0000000..be53e5c
--- /dev/null
+++ b/speech/deepSpeechTest.py
@@ -0,0 +1,20 @@
+from deepspeech import Model
+import scipy.io.wavfile as wav
+import numpy as np
+
+# 鎸囧畾妯″瀷鍜岃瘎鍒嗗櫒鏂囦欢璺緞
+model_path = 'model/deepspeech-0.9.3-models-zh-CN.pbmm'
+scorer_path = 'model/deepspeech-0.9.3-models-zh-CN.scorer'
+
+# 鍒濆鍖朌eepSpeech妯″瀷
+ds = Model(model_path)
+ds.enableExternalScorer(scorer_path)
+
+# 鍔犺浇闊抽鏂囦欢
+audio_path = 'audio/input.wav'
+fs, audio = wav.read(audio_path)
+
+# 杩涜璇煶杞枃瀛�
+text = ds.stt(audio)
+
+print('Transcribed text:', text)
\ No newline at end of file
diff --git a/speech/formatmp3.py b/speech/formatmp3.py
new file mode 100644
index 0000000..2c94476
--- /dev/null
+++ b/speech/formatmp3.py
@@ -0,0 +1,5 @@
+from pydub import AudioSegment
+
+# 鍔犺浇mp3鏂囦欢
+mp3_audio = AudioSegment.from_mp3("input.mp3")
+
diff --git a/speech/whisper2.py b/speech/whisper2.py
new file mode 100644
index 0000000..5ae37cf
--- /dev/null
+++ b/speech/whisper2.py
@@ -0,0 +1 @@
+from faster import WhisperModel
\ No newline at end of file
diff --git a/speech/whisperTest.py b/speech/whisperTest.py
new file mode 100644
index 0000000..7869783
--- /dev/null
+++ b/speech/whisperTest.py
@@ -0,0 +1,10 @@
+import whisper
+model = whisper.load_model("medium")
+result = model.transcribe("audio/input.wav")
+
+print(result["text"])
+
+# tiny 72.1M 鎴戣闂ぇ瀹惰繖涓瘽閲岀粰鏀惧湪鍓嶅彴浜嗘槸鍗＄綏鍓嶅彴鏄悧瀵瑰ソ鐨勮阿璋�
+# base 1XXM 鎴戝竻鍝ラ偅涓彂鍔涚粰浠栨斁鍦ㄥ墠鍙颁簡鏄�7鍙蜂簡鍓嶅彴鏄惂瀵瑰ソ鐨勮阿璋㈠ソ
+# small 461M 鍠備綘濂藉摝鎴戦爢棰ㄧ殑涓�鍊嬬暙鍦扮郸鏀惧湪鍓嶅彴浜嗘槸7铏熸〒鍓嶅彴鏄惂灏嶅ソ鐨勮瑵璎�
+# medium 1.42G 鍠備綘濂戒綘濂芥垜韬悗閭ｅ効鏈変釜娉曞緥缁欐斁鍦ㄥ墠鍙颁簡鏄�7鍙锋ゼ鍓嶅彴鏄惂瀵瑰ソ鐨勮阿璋㈠ソ
\ No newline at end of file
--
Gitblit v1.9.3