except:
continue
classifications_per_box = [0 for _ in range(len(boxes))]
image_detection = np.copy(image)
###Preprocessing of whole image
for step in detection_pipeline:
func = step["function"]
func = globals()[func]
kwargs = step["kwargs"]
image_detection = func(image_detection, **kwargs)
boxs = reader.detect(image_detection)
image_res = np.copy(image_detection)
resultsText += file + "\n"
results_json = {"filename": file, "boxes": []}
fps = 0
tps = 0
fns = 0
ccs = 0
empty = np.zeros_like(image_res)
for i, box in enumerate(boxs[0]):
gt_count += len(boxes)
classifications_per_box = [0 for _ in range(len(boxes))]
image_detection = np.copy(image)
for step in detection_pipeline:
func = step["function"]
func = globals()[func]
kwargs = step["kwargs"]
image_detection = func(image_detection, **kwargs)
image_detection = image_detection.astype('uint8')
boxes_detected = reader.detect(image_detection)
image_res = np.copy(image_detection)
resultsText += file + "\n"
results_json = {"filename": file, "boxes": []}
fps = 0
tps = 0
fns = 0
ccs = 0
empty = np.zeros_like(image_res)
IoU_threshold = 0.5