IOU code

intersection-over-union/Intersection_over_unionII.py

@@ -0,0 +1,231 @@
+# USAGE
+# python intersection_over_union.py
+
+# import the necessary packages
+from collections import namedtuple
+import numpy as np
+import cv2
+import scipy.io
+import yaml
+import sys
+
+# define the `Detection` object
+Detection = namedtuple("Detection", ["image_path", "gt", "pred"])
+
+
+def bb_intersection_over_union(boxA, boxB):
+    # determine the (x, y)-coordinates of the intersection rectangle
+    xA = max(boxA[0], boxB[0])
+    yA = max(boxA[1], boxB[1])
+    xB = min(boxA[2], boxB[2])
+    yB = min(boxA[3], boxB[3])
+
+    # compute the area of intersection rectangle
+    interArea = max(0, xB - xA + 1) * max(0, yB - yA + 1)
+
+    # compute the area of both the prediction and ground-truth
+    # rectangles
+    boxAArea = (boxA[2] - boxA[0] + 1) * (boxA[3] - boxA[1] + 1)
+    boxBArea = (boxB[2] - boxB[0] + 1) * (boxB[3] - boxB[1] + 1)
+
+    # compute the intersection over union by taking the intersection
+    # area and dividing it by the sum of prediction + ground-truth
+    # areas - the interesection area
+    iou = interArea / float(boxAArea + boxBArea - interArea)
+
+    # return the intersection over union value
+    return iou
+
+
+# define the list of example detections
+# examples = [
+#	Detection("image_0002.jpg", [39, 63, 203, 112], [54, 66, 198, 114]),
+#	Detection("image_0016.jpg", [49, 75, 203, 125], [42, 78, 186, 126]),
+#	Detection("image_0075.jpg", [31, 69, 201, 125], [18, 63, 235, 135]),
+#	Detection("image_0090.jpg", [50, 72, 197, 121], [54, 72, 198, 120]),
+#	Detection("image_0120.jpg", [35, 51, 196, 110], [36, 60, 180, 108])]
+# read .mat is impossible
+# detection= scipy.io.loadmat('/home/carlos/Escritorio/dataset/pasillo/epfl_rgbd_pedestrians/epfl_lab/ground_truth_image_plane.mat')
+# file_GT = open('/home/carlos/Escritorio/dataset/pasillo/epfl_rgbd_pedestrians/epfl_corridor/ground_truth_image_plane.yaml')
+file_GT = open('/home/carlos/Escritorio/dataset/pasillo/epfl_rgbd_pedestrians/epfl_lab/ground_truth_image_plane.yaml')
+file_video = open('/home/carlos/Escritorio/dataset/IOU/intersection-over-union/result.yaml')
+
+file_dict_GT = yaml.load(file_GT, Loader=yaml.FullLoader)
+file_dict_video = yaml.load(file_video, Loader=yaml.FullLoader)
+# frames_dict_GT=file_dict_GT['20141008_141323_00']
+frames_dict_GT = file_dict_GT['20140804_160621_00']
+frames_dict_video = file_dict_video['lab_video']
+
+initial = 0  # next(iter(frames_dict_GT))
+
+# extraccion del video
+camera_width = 512
+camera_height = 424
+vidfps = 25
+number_frame = 0
+PATH = '/home/carlos/Escritorio/dataset/pasillo/epfl_rgbd_pedestrians/epfl_lab/20140804_160621_00/out_epfl_rgbd.mp4'
+# PATH='/home/carlos/Escritorio/dataset/pasillo/epfl_rgbd_pedestrians/epfl_corridor/20141008_141323_00/video.mp4'
+cam = cv2.VideoCapture(PATH)
+if cam.isOpened() != True:
+    print("USB Camera Open Error!!!")
+    sys.exit(0)
+
+cam.set(cv2.CAP_PROP_FPS, vidfps)
+cam.set(cv2.CAP_PROP_FRAME_WIDTH, camera_width)
+cam.set(cv2.CAP_PROP_FRAME_HEIGHT, camera_height)
+framenumber = 0
+
+num_of_frame = int(cam.get(cv2.CAP_PROP_FRAME_COUNT))
+cv2.namedWindow("USB_camera", cv2.WINDOW_AUTOSIZE)
+dict_IOU = []
+list_IOU = []
+# read .yaml from dictionary
+
+estado = []
+initial = (next(iter(frames_dict_GT)))
+# loop over the example detections
+while (initial < (num_of_frame - (next(iter(frames_dict_GT))))):
+
+    s, image = cam.read()
+    flag_image = image
+
+    if initial >= (next(iter(frames_dict_GT))):
+
+        sel_GT = frames_dict_GT[initial]
+        sel_video = frames_dict_video[initial + 13]
+        k = 0
+
+        list_IOU_0 = [0, 0, 0, 0, 0, 0, 0]
+        list_IOU_1 = [0, 0, 0, 0, 0, 0, 0]
+        list_IOU_2 = [0, 0, 0, 0, 0, 0, 0]
+        list_IOU_3 = [0, 0, 0, 0, 0, 0, 0]
+        estado = [0, 0, 0, 0]
+
+        for l in list(sel_GT.keys()):
+            # k=next(iter(sel_GT))+l
+
+            # image=flag_image
+            box_gT_dict = sel_GT[l]
+            box_gT_list = list(box_gT_dict)
+            box_gt_func = [box_gT_list[0], box_gT_list[1], box_gT_list[0] + box_gT_list[2],
+                           box_gT_list[1] + box_gT_list[3]]
+
+            for i in list(sel_video.keys()):
+
+                image = flag_image
+                image = cv2.putText(image, str(initial), (50, 50), cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 0, 255), 1)
+                cv2.circle(image, (0, 0), 2, (0, 0, 255), 2)
+                cv2.circle(image, (256, 212), 2, (0, 0, 255), 2)
+                cv2.circle(image, (box_gT_list[0], box_gT_list[1]), 2, (0, 0, 255), 2)
+                image = cv2.rectangle(image, (box_gT_list[0], box_gT_list[1]),
+                                      (box_gT_list[0] + box_gT_list[2], box_gT_list[1] + box_gT_list[3]),
+                                      (0, 250, (l * 50)), 2)
+                box_video_dict = sel_video[i]
+                box_video_list = list(box_video_dict)
+
+                # color BGR
+
+                image = cv2.rectangle(image, (box_video_list[0], box_video_list[1]),
+                                      (box_video_list[2], box_video_list[3]), (255, 0, (i * 50)), 2)
+
+                if l == 0:
+                    list_IOU_0[i] = bb_intersection_over_union(box_gt_func, box_video_list)
+                    estado[l] = 1
+                elif l == 1:
+                    list_IOU_1[i] = bb_intersection_over_union(box_gt_func, box_video_list)
+                    estado[l] = 1
+                elif l == 2:
+                    list_IOU_2[i] = bb_intersection_over_union(box_gt_func, box_video_list)
+                    estado[l] = 1
+                elif l == 3:
+                    list_IOU_3[i] = bb_intersection_over_union(box_gt_func, box_video_list)
+                    estado[l] = 1
+
+        """""cv2.imshow("USB_camera", image)
+        key = cv2.waitKey(0)
+        while key not in [ord('q'), ord('k')]:
+            key = cv2.waitKey(0) """""
+
+        if initial == 34:
+            dict_IOU = {initial: {0: list_IOU_0, 1: list_IOU_1, 2: list_IOU_2, 3: list_IOU_3, "estado": estado}}
+        elif initial > 34:
+            dict_IOU.update({initial: {0: list_IOU_0, 1: list_IOU_1, 2: list_IOU_2, 3: list_IOU_3, "estado": estado}})
+
+        # funcion IOU
+
+    initial += 1
+    # BGR
+# Procesado del diccionarion
+print("Procesando de estadisticas")
+FP_list = []
+TP_list = []
+FN_list = []
+TN_list = []
+
+FP = 0
+TP = 0
+FN = 0
+TN = 0
+dict_resutl = []
+
+# comprobar la IOU
+n = (next(iter(frames_dict_GT)))
+# first input
+FP_dic = {n: {"FP0": 0, "FP1": 0, "FP2": 0, "FP3": 0}}
+TP_dic = {n: {"TP0": 0, "TP1": 0, "TP2": 0, "TP3": 0}}
+FN_dic = {n: {"FN0": 0, "FN1": 0, "FN2": 0, "FN3": 0}}
+TN_dic = {n: {"TN0": 0, "TN1": 0, "TN2": 0, "TN3": 0}}
+while n < (num_of_frame - (next(iter(frames_dict_GT)))):
+
+    if n >= (next(iter(frames_dict_GT))):
+
+        #print("frame numero: ", n)
+        frame_information = dict_IOU.get(n)
+        personas_por_GT = 0
+        FP_list = [0,0,0,0]
+        TP_list = [0,0,0,0]
+        FN_list = [0,0,0,0]
+        TN_list = [0,0,0,0]
+        for i in range(len(frame_information['estado'])):
+
+
+            if frame_information['estado'][i] == 1:
+                personas_por_GT += 1
+                identificador_persona_GT = i
+                if len(set(frame_information[i])) == 1:
+                    FN += 1
+                else:
+                    number_positive = 0
+                    down_umbral = 0
+                    for l in range(len(frame_information[i])):
+                        if frame_information[i][l] >= 0.18:
+                            number_positive += 1
+                        elif 0 < frame_information[i][l] < 0.18:
+                            down_umbral += 1
+
+                    if number_positive >= 1:
+                        FP += (number_positive - 1)
+                        TP += 1
+                    if down_umbral >= 1:
+                        # FN+=1
+                        FN = FN
+            else:
+
+                if len(set(frame_information[i])) == 1:
+
+                    TN += 1
+                else:
+                    FP += 1
+            FP_list[i]=FP
+            TP_list[i]=TP
+            FN_list[i]=FN
+            TN_list[i]=TN
+
+    FP_dic.update({n: {"FP0": FP_list[0], "FP1": FP_list[1], "FP2": FP_list[2], "FP3": FP_list[3]}})
+    TP_dic.update({n: {"TP0": TP_list[0], "TP1": TP_list[1], "TP2": TP_list[2], "TP3": TP_list[3]}})
+    FN_dic.update({n: {"FN0": FN_list[0], "FN1": FN_list[1], "FN2": FN_list[2], "FN3": FN_list[3]}})
+    TN_dic.update({n: {"TN0": TN_list[0], "TN1": TN_list[1], "TN2": TN_list[2], "TN3": TN_list[3]}})
+    n+=1
+
+
+print("hola")

intersection-over-union/intersection_over_union.py

@@ -0,0 +1,119 @@
+# USAGE
+# python intersection_over_union.py
+
+# import the necessary packages
+from collections import namedtuple
+import numpy as np
+import cv2
+import scipy.io
+import yaml
+import sys
+
+# define the `Detection` object
+Detection = namedtuple("Detection", ["image_path", "gt", "pred"])
+
+def bb_intersection_over_union(boxA, boxB):
+	# determine the (x, y)-coordinates of the intersection rectangle
+	xA = max(boxA[0], boxB[0])
+	yA = max(boxA[1], boxB[1])
+	xB = min(boxA[2], boxB[2])
+	yB = min(boxA[3], boxB[3])
+
+	# compute the area of intersection rectangle
+	interArea = max(0, xB - xA + 1) * max(0, yB - yA + 1)
+
+	# compute the area of both the prediction and ground-truth
+	# rectangles
+	boxAArea = (boxA[2] - boxA[0] + 1) * (boxA[3] - boxA[1] + 1)
+	boxBArea = (boxB[2] - boxB[0] + 1) * (boxB[3] - boxB[1] + 1)
+
+	# compute the intersection over union by taking the intersection
+	# area and dividing it by the sum of prediction + ground-truth
+	# areas - the interesection area
+	iou = interArea / float(boxAArea + boxBArea - interArea)
+
+	# return the intersection over union value
+	return iou
+
+# define the list of example detections
+#examples = [
+#	Detection("image_0002.jpg", [39, 63, 203, 112], [54, 66, 198, 114]),
+#	Detection("image_0016.jpg", [49, 75, 203, 125], [42, 78, 186, 126]),
+#	Detection("image_0075.jpg", [31, 69, 201, 125], [18, 63, 235, 135]),
+#	Detection("image_0090.jpg", [50, 72, 197, 121], [54, 72, 198, 120]),
+#	Detection("image_0120.jpg", [35, 51, 196, 110], [36, 60, 180, 108])]
+#read .mat is impossible
+#detection= scipy.io.loadmat('/home/carlos/Escritorio/dataset/pasillo/epfl_rgbd_pedestrians/epfl_lab/ground_truth_image_plane.mat')
+
+file_GT = open('/home/carlos/Escritorio/dataset/pasillo/epfl_rgbd_pedestrians/epfl_lab/ground_truth_image_plane.yaml')
+file_video= open('/home/carlos/Escritorio/dataset/IOU/intersection-over-union/result.yaml')
+
+file_dict_GT=yaml.load(file_GT,Loader=yaml.FullLoader)
+file_dict_video= yaml.load(file_video,Loader=yaml.FullLoader)
+
+frames_dict_GT=file_dict_GT['20140804_160621_00']
+frames_dict_video=file_dict_video['lab_video']
+
+initial=0#next(iter(frames_dict_GT))
+
+#extraccion del video
+camera_width = 640
+camera_height = 480
+vidfps = 25
+number_frame=0
+PATH='/home/carlos/Escritorio/dataset/pasillo/epfl_rgbd_pedestrians/epfl_lab/20140804_160621_00/out_epfl_rgbd.mp4'
+cam = cv2.VideoCapture(PATH)
+if cam.isOpened() != True:
+	print("USB Camera Open Error!!!")
+	sys.exit(0)
+
+cam.set(cv2.CAP_PROP_FPS, vidfps)
+cam.set(cv2.CAP_PROP_FRAME_WIDTH, camera_width)
+cam.set(cv2.CAP_PROP_FRAME_HEIGHT, camera_height)
+framenumber = 0
+
+num_of_frame = int(cam.get(cv2.CAP_PROP_FRAME_COUNT))
+cv2.namedWindow("USB_camera", cv2.WINDOW_AUTOSIZE)
+
+
+#read .yaml from dictionary
+
+
+# loop over the example detections
+while(1):
+
+	s, image = cam.read()
+
+	if initial >= (next(iter(frames_dict_GT))):
+
+		sel_GT = frames_dict_GT[initial]
+		sel_video = frames_dict_video[initial+13]
+		k=0
+		for l in list(sel_GT.keys()):
+			#k=next(iter(sel_GT))+l
+
+			box_gT_dict = sel_GT[l]
+			box_gT_list = list(box_gT_dict)
+			image = cv2.rectangle(image, (box_gT_list[0], box_gT_list[1]), (box_gT_list[0] + box_gT_list[2], box_gT_list[1] + box_gT_list[3]), (255,0, (l*50)), 2)
+
+
+		for i in list(sel_video.keys()):
+			box_video_dict=sel_video[i]
+			box_video_list=list(box_video_dict)
+
+			#color BGR
+
+			image = cv2.rectangle(image,(box_video_list[0],box_video_list[1]),(box_video_list[2],box_video_list[3]),(255,0, (i*50)),2)
+
+			#funcion IOU
+
+	initial+=1
+	#BGR
+	cv2.imshow("USB_camera", image)
+	key = cv2.waitKey(0)
+	while key not in [ord('q'), ord('k')]:
+		key = cv2.waitKey(0)
+
+	#comprobar la IOU
+
+