icepatchnet 0.1.0__tar.gz

icepatchnet-0.1.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Sai Vikas Amaraneni
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

icepatchnet-0.1.0/PKG-INFO

@@ -0,0 +1,60 @@
+Metadata-Version: 2.4
+Name: icepatchnet
+Version: 0.1.0
+Summary: Ice-PatchNet: A patch (segmented images) based CNN workflow
+Author: Sai Vikas Amaraneni, Maloy Kumar Devnath, Shalini Srinivasa, Chhaya Kulkarni, Sudip Chakraborty, Vandana P. Janeja
+License: iHARP
+Project-URL: Homepage, https://github.com/vikasamaraneni333/icepatchnet
+Project-URL: Source, https://github.com/vikasamaraneni333/icepatchnet
+Project-URL: Issues, https://github.com/vikasamaraneni333/icepatchnet/issues
+Keywords: icepatchnet,Ice-PatchNet,patch based CNN,Convolutional Neural Network,segmentation of images,image prediction,spatio temporal data mining,geospatial,geoai,climate,cryosphere,machine learning,data science
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Science/Research
+Classifier: Topic :: Scientific/Engineering
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Classifier: Topic :: Scientific/Engineering :: Information Analysis
+Classifier: Topic :: Scientific/Engineering :: GIS
+Classifier: Topic :: Scientific/Engineering :: Atmospheric Science
+Classifier: Topic :: Software Development :: Libraries :: Python Modules
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3 :: Only
+Classifier: Operating System :: OS Independent
+Requires-Python: >=3.8
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: numpy
+Requires-Dist: torch
+Requires-Dist: opencv-python
+Requires-Dist: matplotlib
+Requires-Dist: tqdm
+Dynamic: license-file
+
+# Predicting Antarctic Sea Ice with Scalable Deep Learning
+## Ice-PatchNet
+Ice-PatchNet is a novel, scalable deep learning framework designed to predict daily Sea Ice Extent (SIE) over the Antarctic region. By using a patch-based segmentation approach, the model can efficiently capture localized spatiotemporal features.
+
+## Performance and Scope
+
+- Target Region: Optimized for the Antarctic and Arctic polar regions.
+- Resolution: Handles satellite imagery at a 25km spatial resolution.
+- Predictive Windows: Supports lead-time predictions for 1-day, 7-day, and 14-day horizons.
+
+### Important Version info
+This version of Ice-PatchNet has only been tested on the images of sea ice extent over the Arctic region and the Antarctic region.
+
+----------------------------------------------------------------------------
+
+## Method Overview
+
+1. **Image Preprocessing:** High-resolution SIE images are converted to grayscale to reduce complexity and computational overhead.
+2. **Patch Segmentation:** The images are divided into a user-defined number of non-overlapping patches. This enables an equal distribution of SIE features and improves the model's ability to distinguish between land, open water, and ice.
+3. **Spatiotemporal Feature Extraction:** Each patch is treated as a single unit and is processed through a deep convolutional neural network (3 layers, $3 \times 3$ kernels) with ReLU activation to identify localized patterns.
+4. **Patches Reassembly:** Predicted patches are re-aligned using a global counter to reconstruct the full geographic extent without spatial distortion.
+
+------------------------
+
+## Citation
+If you utilize this framework or the patch-based methodology in your research, please cite our ICDM '25 paper:
+```
+Amaraneni, S. V., Devnath, M. K., Srinivasa, S., Kulkarni, C., Chakraborty, S., & Janeja, V. P. (2025). Predicting Antarctic sea ice with scalable deep learning models. In Proceedings of the IEEE ICDM Workshop on Spatial and Spatiotemporal Data Mining (SSTDM).
+```

icepatchnet-0.1.0/README.md

@@ -0,0 +1,29 @@
+# Predicting Antarctic Sea Ice with Scalable Deep Learning
+## Ice-PatchNet
+Ice-PatchNet is a novel, scalable deep learning framework designed to predict daily Sea Ice Extent (SIE) over the Antarctic region. By using a patch-based segmentation approach, the model can efficiently capture localized spatiotemporal features.
+
+## Performance and Scope
+
+- Target Region: Optimized for the Antarctic and Arctic polar regions.
+- Resolution: Handles satellite imagery at a 25km spatial resolution.
+- Predictive Windows: Supports lead-time predictions for 1-day, 7-day, and 14-day horizons.
+
+### Important Version info
+This version of Ice-PatchNet has only been tested on the images of sea ice extent over the Arctic region and the Antarctic region.
+
+----------------------------------------------------------------------------
+
+## Method Overview
+
+1. **Image Preprocessing:** High-resolution SIE images are converted to grayscale to reduce complexity and computational overhead.
+2. **Patch Segmentation:** The images are divided into a user-defined number of non-overlapping patches. This enables an equal distribution of SIE features and improves the model's ability to distinguish between land, open water, and ice.
+3. **Spatiotemporal Feature Extraction:** Each patch is treated as a single unit and is processed through a deep convolutional neural network (3 layers, $3 \times 3$ kernels) with ReLU activation to identify localized patterns.
+4. **Patches Reassembly:** Predicted patches are re-aligned using a global counter to reconstruct the full geographic extent without spatial distortion.
+
+------------------------
+
+## Citation
+If you utilize this framework or the patch-based methodology in your research, please cite our ICDM '25 paper:
+```
+Amaraneni, S. V., Devnath, M. K., Srinivasa, S., Kulkarni, C., Chakraborty, S., & Janeja, V. P. (2025). Predicting Antarctic sea ice with scalable deep learning models. In Proceedings of the IEEE ICDM Workshop on Spatial and Spatiotemporal Data Mining (SSTDM).
+```

icepatchnet-0.1.0/pyproject.toml

@@ -0,0 +1,117 @@
+# =========================================================
+# Build system configuration
+# =========================================================
+[build-system]
+requires = ["setuptools>=61.0"]
+build-backend = "setuptools.build_meta"
+
+
+# =========================================================
+# Project metadata (MOST IMPORTANT SECTION)
+# =========================================================
+[project]
+
+# ⚠️ Provide your PACKAGE Name
+name = "icepatchnet"
+version = "0.1.0"
+
+# 🔁 Update according your project scope or description
+description = "Ice-PatchNet: A patch (segmented images) based CNN workflow"
+
+# ⚠️ Write the decription how you want to represent your Packages
+readme = "README.md"
+
+# ⚠️ Change only if dropping support for older Python
+requires-python = ">=3.8"
+
+# =========================================================
+# Authors (update if contributors change)
+# =========================================================
+authors = [
+  { name = "Sai Vikas Amaraneni" },
+  { name = "Maloy Kumar Devnath" },
+  { name = "Shalini Srinivasa" },
+  { name = "Chhaya Kulkarni" },
+  { name = "Sudip Chakraborty" },
+  { name = "Vandana P. Janeja" }
+]
+
+# =========================================================
+# License information
+# ---------------------------------------------------------
+# ⚠️ PyPI prefers SPDX identifiers (MIT, BSD-3-Clause, etc.)
+# If you later change license → UPDATE THIS
+# =========================================================
+license = { text = "iHARP" }
+
+# =========================================================
+# Runtime dependencies
+# =========================================================
+dependencies = [
+  "numpy",
+  "torch",
+  "opencv-python",
+  "matplotlib",
+  "tqdm"
+]
+
+# =========================================================
+# Keywords (PyPI discovery & search)
+# =========================================================
+keywords = [
+  "icepatchnet",
+  "Ice-PatchNet",
+  "patch based CNN",
+  "Convolutional Neural Network",
+  "segmentation of images",
+  "image prediction",
+  "spatio temporal data mining",
+  "geospatial",
+  "geoai",
+  "climate",
+  "cryosphere",
+  "machine learning",
+  "data science"
+]
+
+# =========================================================
+# Classifiers (MUST be valid PyPI classifiers). Update for your packages
+# =========================================================
+classifiers = [
+  "Development Status :: 3 - Alpha",
+  "Intended Audience :: Science/Research",
+
+  "Topic :: Scientific/Engineering",
+  "Topic :: Scientific/Engineering :: Artificial Intelligence",
+  "Topic :: Scientific/Engineering :: Information Analysis",
+  "Topic :: Scientific/Engineering :: GIS",
+  "Topic :: Scientific/Engineering :: Atmospheric Science",
+
+  "Topic :: Software Development :: Libraries :: Python Modules",
+
+  "Programming Language :: Python :: 3",
+  "Programming Language :: Python :: 3 :: Only",
+
+  "Operating System :: OS Independent"
+]
+
+# =========================================================
+# Project URLs
+# =========================================================
+[project.urls]
+Homepage = "https://github.com/vikasamaraneni333/icepatchnet"
+Source   = "https://github.com/vikasamaraneni333/icepatchnet"
+Issues   = "https://github.com/vikasamaraneni333/icepatchnet/issues"
+
+# =========================================================
+# Setuptools configuration
+# =========================================================
+[tool.setuptools]
+package-dir = { "" = "src" }
+
+
+# =========================================================
+# Package discovery
+# =========================================================
+[tool.setuptools.packages.find]
+where = ["src"]

icepatchnet-0.1.0/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+

icepatchnet-0.1.0/src/icepatchnet/__init__.py

@@ -0,0 +1,4 @@
+from .utils import to_grayscale, extract_patches, reconstruct_image
+from .model import CNNModel
+
+__all__ = ["to_grayscale", "extract_patches", "reconstruct_image", 'CNNModel']

icepatchnet-0.1.0/src/icepatchnet/model.py

@@ -0,0 +1,113 @@
+import torch
+import torch.nn as nn
+
+class CNNModel(nn.Module):
+    """
+    A configurable patch-based CNN for Sea Ice Extent prediction.
+
+    Parameters
+    ----------
+    num_layers : int, optional
+        Number of convolutional layers. Default is 3.
+    kernel_size : int or list of int, optional
+        Kernel size for each layer.
+        - If a single int  → same kernel used for ALL layers.
+        - If a list of int → one kernel size per layer (length must match num_layers).
+        Default is 3 (i.e., 3×3 kernels on all layers).
+    base_filters : int, optional
+        Number of filters in the first hidden layer. Each subsequent hidden
+        layer doubles the filters. Default is 64.
+
+    Examples
+    --------
+    # Default (matches original model — 3 layers, 3×3 kernels)
+    model = CNNModel()
+
+    # 5 layers, all with 3×3 kernels
+    model = CNNModel(num_layers=5)
+
+    # 4 layers, all with 5×5 kernels
+    model = CNNModel(num_layers=4, kernel_size=5)
+
+    # 4 layers, each with a different kernel size
+    model = CNNModel(num_layers=4, kernel_size=[3, 5, 5, 3])
+    """
+
+    def __init__(self, num_layers=3, kernel_size=3, base_filters=64):
+
+        super(CNNModel, self).__init__()
+
+        # ── Validate num_layers ──────────────────────────────────────
+        if not isinstance(num_layers, int) or num_layers < 1:
+            raise ValueError(f"`num_layers` must be a positive integer, got {num_layers!r}")
+
+        # ── Resolve kernel sizes ─────────────────────────────────────
+        if isinstance(kernel_size, int):
+            # Same kernel for every layer
+            kernel_sizes = [kernel_size] * num_layers
+        elif isinstance(kernel_size, (list, tuple)):
+            if len(kernel_size) != num_layers:
+                raise ValueError(
+                    f"`kernel_size` list length ({len(kernel_size)}) "
+                    f"must match `num_layers` ({num_layers})"
+                )
+            kernel_sizes = list(kernel_size)
+        else:
+            raise TypeError(
+                f"`kernel_size` must be an int or list of ints, got {type(kernel_size).__name__!r}"
+            )
+
+        # Validate each individual kernel size (must be positive odd integer)
+        for i, k in enumerate(kernel_sizes):
+            if not isinstance(k, int) or k < 1 or k % 2 == 0:
+                raise ValueError(
+                    f"Each kernel size must be a positive odd integer. "
+                    f"Got kernel_sizes[{i}] = {k!r}"
+                )
+
+        # ── Build the layer stack ────────────────────────────────────
+        # Channel progression:
+        #   Layer 1      : 1                              → base_filters
+        #   Layer 2..N-1 : base_filters * 2^(i-1)        → base_filters * 2^i
+        #   Last layer   : <previous out_channels>        → 1  (single channel output)
+
+        layers = []
+        in_channels = 1
+        out_channels = base_filters
+
+        for i in range(num_layers):
+            is_last = (i == num_layers - 1)
+            k = kernel_sizes[i]
+            padding = k // 2   # 'same' padding — preserves H and W dimensions
+
+            if is_last:
+                out_channels = 1
+            
+            layers.append(nn.Conv2d(in_channels, out_channels, kernel_size=k, padding=padding))
+            layers.append(nn.ReLU())
+
+            in_channels = out_channels
+            if not is_last:
+                out_channels = out_channels * 2   # double filters for next layer
+
+        self.network = nn.Sequential(*layers)
+
+        # Store config for easy inspection
+        self.num_layers   = num_layers
+        self.kernel_sizes = kernel_sizes
+        self.base_filters = base_filters
+
+    def forward(self, x):
+        return self.network(x)
+
+    def __repr__(self):
+        lines = [
+            f"CNNModel(",
+            f"  num_layers  = {self.num_layers}",
+            f"  kernel_sizes= {self.kernel_sizes}",
+            f"  base_filters= {self.base_filters}",
+        ]
+        for name, module in self.network.named_children():
+            lines.append(f"  ({name}): {module}")
+        lines.append(")")
+        return "\n".join(lines)

icepatchnet-0.1.0/src/icepatchnet/utils.py

@@ -0,0 +1,75 @@
+import numpy as np
+import math
+import cv2
+
+def to_grayscale(filelist):
+    rgb_image_list = []
+    gray_image_list = []
+
+    for file in filelist:
+        image_rgb = cv2.imread(file)
+        rgb_image_list.append(image_rgb)
+
+        gray_image = cv2.cvtColor(image_rgb, cv2.COLOR_BGR2GRAY)
+        gray_image_list.append(gray_image)
+
+    return gray_image_list,rgb_image_list
+
+def extract_patches(images, num_patches):
+    all_patches = []
+
+    for image in images:
+        # Calculate the patch size
+        image_height, image_width = image.shape
+        patch_height = math.floor(image_height / int(math.sqrt(num_patches)))
+        patch_width = math.floor(image_width / int(math.sqrt(num_patches)))
+        patch_size = (patch_width, patch_height)
+
+        # Extract patches from the image
+        patches = []
+        row_step = image_height // int(math.sqrt(num_patches))
+        col_step = image_width // int(math.sqrt(num_patches))
+
+        for row in range(0, image_height - patch_height + 1, row_step):
+            for col in range(0, image_width - patch_width + 1, col_step):
+                patch = image[row:row+patch_height, col:col+patch_width]
+                patches.append(patch)
+
+        all_patches.append(patches)
+
+    return np.array(all_patches)
+
+def reconstruct_image(processed_patches, original_shape, num_images):
+    original_height, original_width = original_shape
+
+    # Get the patch height and patch width from the processed_patches array
+    _, channels, patch_height, patch_width = processed_patches.shape
+
+    # Number of patches along each dimension
+    num_patches_height = original_height // patch_height
+    num_patches_width = original_width // patch_width
+
+    # Create an empty array to store the reconstructed images
+    reconstructed_images = np.zeros((num_images, original_height, original_width), dtype=np.uint8)
+
+    # Iterate over the images
+    patch_index = 0
+    for img_idx in range(num_images):
+        # Iterate over the patches and place them in the reconstructed image
+        for row in range(num_patches_height):
+            for col in range(num_patches_width):
+                # Check if the patch index is within the valid range
+                if patch_index >= processed_patches.shape[0]:
+                    break
+
+                # Calculate the position of the current patch in the reconstructed image
+                row_start = row * patch_height
+                row_end = row_start + patch_height
+                col_start = col * patch_width
+                col_end = col_start + patch_width
+
+                # Copy the processed patch into the reconstructed image
+                reconstructed_images[img_idx, row_start:row_end, col_start:col_end] = processed_patches[patch_index, 0]
+                patch_index += 1
+
+    return reconstructed_images

icepatchnet-0.1.0/src/icepatchnet.egg-info/PKG-INFO

@@ -0,0 +1,60 @@
+Metadata-Version: 2.4
+Name: icepatchnet
+Version: 0.1.0
+Summary: Ice-PatchNet: A patch (segmented images) based CNN workflow
+Author: Sai Vikas Amaraneni, Maloy Kumar Devnath, Shalini Srinivasa, Chhaya Kulkarni, Sudip Chakraborty, Vandana P. Janeja
+License: iHARP
+Project-URL: Homepage, https://github.com/vikasamaraneni333/icepatchnet
+Project-URL: Source, https://github.com/vikasamaraneni333/icepatchnet
+Project-URL: Issues, https://github.com/vikasamaraneni333/icepatchnet/issues
+Keywords: icepatchnet,Ice-PatchNet,patch based CNN,Convolutional Neural Network,segmentation of images,image prediction,spatio temporal data mining,geospatial,geoai,climate,cryosphere,machine learning,data science
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Science/Research
+Classifier: Topic :: Scientific/Engineering
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Classifier: Topic :: Scientific/Engineering :: Information Analysis
+Classifier: Topic :: Scientific/Engineering :: GIS
+Classifier: Topic :: Scientific/Engineering :: Atmospheric Science
+Classifier: Topic :: Software Development :: Libraries :: Python Modules
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3 :: Only
+Classifier: Operating System :: OS Independent
+Requires-Python: >=3.8
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: numpy
+Requires-Dist: torch
+Requires-Dist: opencv-python
+Requires-Dist: matplotlib
+Requires-Dist: tqdm
+Dynamic: license-file
+
+# Predicting Antarctic Sea Ice with Scalable Deep Learning
+## Ice-PatchNet
+Ice-PatchNet is a novel, scalable deep learning framework designed to predict daily Sea Ice Extent (SIE) over the Antarctic region. By using a patch-based segmentation approach, the model can efficiently capture localized spatiotemporal features.
+
+## Performance and Scope
+
+- Target Region: Optimized for the Antarctic and Arctic polar regions.
+- Resolution: Handles satellite imagery at a 25km spatial resolution.
+- Predictive Windows: Supports lead-time predictions for 1-day, 7-day, and 14-day horizons.
+
+### Important Version info
+This version of Ice-PatchNet has only been tested on the images of sea ice extent over the Arctic region and the Antarctic region.
+
+----------------------------------------------------------------------------
+
+## Method Overview
+
+1. **Image Preprocessing:** High-resolution SIE images are converted to grayscale to reduce complexity and computational overhead.
+2. **Patch Segmentation:** The images are divided into a user-defined number of non-overlapping patches. This enables an equal distribution of SIE features and improves the model's ability to distinguish between land, open water, and ice.
+3. **Spatiotemporal Feature Extraction:** Each patch is treated as a single unit and is processed through a deep convolutional neural network (3 layers, $3 \times 3$ kernels) with ReLU activation to identify localized patterns.
+4. **Patches Reassembly:** Predicted patches are re-aligned using a global counter to reconstruct the full geographic extent without spatial distortion.
+
+------------------------
+
+## Citation
+If you utilize this framework or the patch-based methodology in your research, please cite our ICDM '25 paper:
+```
+Amaraneni, S. V., Devnath, M. K., Srinivasa, S., Kulkarni, C., Chakraborty, S., & Janeja, V. P. (2025). Predicting Antarctic sea ice with scalable deep learning models. In Proceedings of the IEEE ICDM Workshop on Spatial and Spatiotemporal Data Mining (SSTDM).
+```

icepatchnet-0.1.0/src/icepatchnet.egg-info/SOURCES.txt

@@ -0,0 +1,11 @@
+LICENSE
+README.md
+pyproject.toml
+src/icepatchnet/__init__.py
+src/icepatchnet/model.py
+src/icepatchnet/utils.py
+src/icepatchnet.egg-info/PKG-INFO
+src/icepatchnet.egg-info/SOURCES.txt
+src/icepatchnet.egg-info/dependency_links.txt
+src/icepatchnet.egg-info/requires.txt
+src/icepatchnet.egg-info/top_level.txt

icepatchnet-0.1.0/src/icepatchnet.egg-info/dependency_links.txt

@@ -0,0 +1 @@
+

icepatchnet-0.1.0/src/icepatchnet.egg-info/requires.txt

@@ -0,0 +1,5 @@
+numpy
+torch
+opencv-python
+matplotlib
+tqdm

icepatchnet-0.1.0/src/icepatchnet.egg-info/top_level.txt

@@ -0,0 +1 @@
+icepatchnet