app.py

import os
import cv2
import shutil
from PIL import Image
import numpy as np
import torch
from torch.autograd import Variable
from torchvision import transforms
import torch.nn.functional as F
from flask import Flask, request, jsonify, render_template, send_from_directory
import warnings
warnings.filterwarnings("ignore")

app = Flask(__name__)

# 一時ファイル保存用ディレクトリ
UPLOAD_FOLDER = 'uploads'
RESULT_FOLDER = 'results'
EXAMPLES_FOLDER = 'examples'
os.makedirs(UPLOAD_FOLDER, exist_ok=True)
os.makedirs(RESULT_FOLDER, exist_ok=True)
os.makedirs(EXAMPLES_FOLDER, exist_ok=True)

# モデル関連のインポートと初期化
def initialize_model():
    # Clean up previous installations
    if os.path.exists("DIS"):
        shutil.rmtree("DIS")
    if os.path.exists("saved_models"):
        shutil.rmtree("saved_models")
    
    # Clone repository and setup model
    os.system("git clone https://github.com/xuebinqin/DIS")
    os.system("mv DIS/IS-Net/* .")
    
    # Import after setup
    from data_loader_cache import normalize, im_reader, im_preprocess 
    from models import ISNetDIS

    device = 'cuda' if torch.cuda.is_available() else 'cpu'

    # Setup model directories
    if not os.path.exists("saved_models"):
        os.mkdir("saved_models")
        os.system("mv isnet.pth saved_models/")
    
    # Set Parameters
    hypar = {
        "model_path": "./saved_models",
        "restore_model": "isnet.pth",
        "interm_sup": False,
        "model_digit": "full",
        "seed": 0,
        "cache_size": [1024, 1024],
        "input_size": [1024, 1024],
        "crop_size": [1024, 1024],
        "model": ISNetDIS()
    }

    # Build Model
    net = build_model(hypar, device)
    return net, hypar, device

class GOSNormalize(object):
    def __init__(self, mean=[0.485,0.456,0.406], std=[0.229,0.224,0.225]):
        self.mean = mean
        self.std = std

    def __call__(self, image):
        image = normalize(image, self.mean, self.std)
        return image

transform = transforms.Compose([GOSNormalize([0.5,0.5,0.5],[1.0,1.0,1.0])])

def load_image(im_path, hypar):
    im = im_reader(im_path)
    im, im_shp = im_preprocess(im, hypar["cache_size"])
    im = torch.divide(im, 255.0)
    shape = torch.from_numpy(np.array(im_shp))
    return transform(im).unsqueeze(0), shape.unsqueeze(0)

def build_model(hypar, device):
    net = hypar["model"]
    
    if hypar["model_digit"] == "half":
        net.half()
        for layer in net.modules():
            if isinstance(layer, nn.BatchNorm2d):
                layer.float()

    net.to(device)

    if hypar["restore_model"] != "":
        net.load_state_dict(torch.load(hypar["model_path"]+"/"+hypar["restore_model"], map_location=device))
        net.to(device)
    net.eval()  
    return net

def predict(net, inputs_val, shapes_val, hypar, device):
    net.eval()

    if hypar["model_digit"] == "full":
        inputs_val = inputs_val.type(torch.FloatTensor)
    else:
        inputs_val = inputs_val.type(torch.HalfTensor)

    inputs_val_v = Variable(inputs_val, requires_grad=False).to(device)
    ds_val = net(inputs_val_v)[0]
    pred_val = ds_val[0][0,:,:,:]

    pred_val = torch.squeeze(F.upsample(torch.unsqueeze(pred_val,0),(shapes_val[0][0],shapes_val[0][1]),mode='bilinear'))

    ma = torch.max(pred_val)
    mi = torch.min(pred_val)
    pred_val = (pred_val-mi)/(ma-mi)

    if device == 'cuda': torch.cuda.empty_cache()
    return (pred_val.detach().cpu().numpy()*255).astype(np.uint8)

@app.route('/')
def index():
    return render_template('index.html')

@app.route('/examples/<filename>')
def serve_example(filename):
    # サンプル画像がなければダウンロード
    example_path = os.path.join(EXAMPLES_FOLDER, filename)
    if not os.path.exists(example_path):
        if filename == 'robot.png':
            os.system(f"wget https://raw.githubusercontent.com/xuebinqin/DIS/main/IS-Net/robot.png -O {example_path}")
        elif filename == 'ship.png':
            os.system(f"wget https://raw.githubusercontent.com/xuebinqin/DIS/main/IS-Net/ship.png -O {example_path}")
    
    return send_from_directory(EXAMPLES_FOLDER, filename)

@app.route('/api/process', methods=['POST'])
def process_image():
    if 'image' not in request.files:
        return jsonify({"error": "No image provided"}), 400
    
    file = request.files['image']
    if file.filename == '':
        return jsonify({"error": "No selected file"}), 400
    
    # 毎回モデルを初期化
    net, hypar, device = initialize_model()
    
    # ファイルを保存
    upload_path = os.path.join(UPLOAD_FOLDER, file.filename)
    file.save(upload_path)
    
    try:
        # 画像処理
        image_tensor, orig_size = load_image(upload_path, hypar) 
        mask = predict(net, image_tensor, orig_size, hypar, device)
        
        # 結果を保存
        original_filename = os.path.splitext(file.filename)[0]
        result_rgba_path = os.path.join(RESULT_FOLDER, f"{original_filename}_rgba.png")
        result_mask_path = os.path.join(RESULT_FOLDER, f"{original_filename}_mask.png")
        
        pil_mask = Image.fromarray(mask).convert('L')
        im_rgb = Image.open(upload_path).convert("RGB")
        im_rgba = im_rgb.copy()
        im_rgba.putalpha(pil_mask)
        
        im_rgba.save(result_rgba_path)
        pil_mask.save(result_mask_path)
        
        # 結果のURLを返す
        return jsonify({
            "original": f"/{UPLOAD_FOLDER}/{file.filename}",
            "rgba": f"/{RESULT_FOLDER}/{original_filename}_rgba.png",
            "mask": f"/{RESULT_FOLDER}/{original_filename}_mask.png",
            "filename": file.filename
        })
    except Exception as e:
        return jsonify({"error": str(e)}), 500

@app.route(f'/{UPLOAD_FOLDER}/<filename>')
def serve_upload(filename):
    return send_from_directory(UPLOAD_FOLDER, filename)

@app.route(f'/{RESULT_FOLDER}/<filename>')
def serve_result(filename):
    return send_from_directory(RESULT_FOLDER, filename)

if __name__ == '__main__':
    app.run(host='0.0.0.0', port=7860, debug=True)