111 lines
4.6 KiB
Python
111 lines
4.6 KiB
Python
"""
|
|
@author: kazeyori
|
|
@title: Quick Image Sequence Process
|
|
@nickname: QuickSeq
|
|
@description: A ComfyUI plugin for efficient image sequence processing. Features frame insertion, duplication, and removal with intuitive controls.
|
|
"""
|
|
|
|
# quick_image_sequence_process.py
|
|
import numpy as np
|
|
import torch
|
|
|
|
class QuickImageSequenceProcessNode:
|
|
def __init__(self):
|
|
pass
|
|
|
|
@classmethod
|
|
def INPUT_TYPES(cls):
|
|
return {
|
|
"required": {
|
|
"images": ("IMAGE",), # Input image sequence
|
|
"frames_between": ("INT", {"default": 1}), # Number of frames to copy between each frame
|
|
"copy_frame": (["previous", "next"],), # Whether to copy the previous or next frame
|
|
"quick_process_first_frames": ("INT", {"default": 0}), # Number of frames to add/remove at the beginning (negative values for removal)
|
|
"quick_process_last_frames": ("INT", {"default": 0}), # Number of frames to add/remove at the end (negative values for removal)
|
|
"process_F_L_frames_first": (["yes", "no"],), # Whether to prioritize processing first and last frames
|
|
},
|
|
}
|
|
|
|
RETURN_TYPES = ("IMAGE", "INT", "INT", "INT") # Output: image, width, height, count
|
|
RETURN_NAMES = ("image", "width", "height", "count")
|
|
FUNCTION = "edit_image_sequence"
|
|
|
|
CATEGORY = "image/sequence"
|
|
|
|
def process_between_frames(self, images, frames_between, copy_frame):
|
|
"""Process middle frames (including frame insertion)"""
|
|
new_images = []
|
|
for i in range(len(images)):
|
|
# Add current frame
|
|
new_images.append(images[i])
|
|
|
|
# If not the last frame, insert copied frames between current and next
|
|
if i < len(images) - 1:
|
|
if copy_frame == "previous":
|
|
frame_to_copy = images[i]
|
|
else:
|
|
frame_to_copy = images[i + 1]
|
|
|
|
for _ in range(frames_between):
|
|
new_images.append(frame_to_copy)
|
|
return new_images
|
|
|
|
def process_end_frames(self, images, quick_process_first_frames, quick_process_last_frames):
|
|
"""Process first and last frames (adding or removing)"""
|
|
new_images = list(images) # Convert to list for modification
|
|
|
|
if quick_process_first_frames >= 0:
|
|
# Add frames at the beginning
|
|
first_frame = new_images[0]
|
|
new_images = [first_frame] * quick_process_first_frames + new_images
|
|
else:
|
|
# Remove frames from the beginning
|
|
new_images = new_images[-quick_process_first_frames:]
|
|
|
|
if quick_process_last_frames >= 0:
|
|
# Add frames at the end
|
|
last_frame = new_images[-1]
|
|
new_images.extend([last_frame] * quick_process_last_frames)
|
|
else:
|
|
# Remove frames from the end
|
|
new_images = new_images[:quick_process_last_frames]
|
|
|
|
return new_images
|
|
|
|
def edit_image_sequence(self, images, frames_between, copy_frame, quick_process_first_frames, quick_process_last_frames, process_F_L_frames_first):
|
|
"""
|
|
Edit image sequence:
|
|
1. Process order determined by process_F_L_frames_first
|
|
2. Insert specified number of copied frames between frames
|
|
3. Add or remove frames at the beginning and end
|
|
"""
|
|
# Ensure input is PyTorch tensor
|
|
if isinstance(images, np.ndarray):
|
|
images = torch.from_numpy(images)
|
|
|
|
if process_F_L_frames_first == "yes":
|
|
# Process first/last frames first, then middle frames
|
|
intermediate_images = self.process_end_frames(images, quick_process_first_frames, quick_process_last_frames)
|
|
new_images = self.process_between_frames(intermediate_images, frames_between, copy_frame)
|
|
else:
|
|
# Process middle frames first, then first/last frames
|
|
intermediate_images = self.process_between_frames(images, frames_between, copy_frame)
|
|
new_images = self.process_end_frames(intermediate_images, quick_process_first_frames, quick_process_last_frames)
|
|
|
|
# Ensure at least one frame remains
|
|
if not new_images:
|
|
raise ValueError("No frames remaining after processing. Please adjust parameters to ensure at least one frame.")
|
|
|
|
# Convert image sequence to PyTorch tensor
|
|
new_images = torch.stack(new_images)
|
|
|
|
# Get dimensions
|
|
width = new_images.shape[2]
|
|
height = new_images.shape[1]
|
|
frame_count = len(new_images)
|
|
|
|
# Display frame count
|
|
print(f"Processed sequence contains {frame_count} frames")
|
|
|
|
# Return results
|
|
return (new_images, width, height, frame_count) |