Initial HF Commit

app.py

@@ -0,0 +1,27 @@
+import gradio as gr
+from utils import process_hsv_image
+
+with gr.Blocks() as app:
+    gr.Markdown("# Color Segmentation Tool using HSV Colorspace")
+    gr.Markdown("Upload an image see the HSV histogram and adjust the Hue bounds to obtain segmented Image based on the range.")
+
+    with gr.Row():
+        with gr.Column():
+            gr.Markdown("### Input Controls")
+            image_input = gr.Image(type="numpy", label="Upload Image") 
+            lower = gr.Slider(minimum=0, maximum=180, label="Lower Bound Hue Value", value=0)
+            upper = gr.Slider(minimum=0, maximum=180, label="Upper Bound Hue Value", value=180)
+            update_button = gr.Button("Update", variant="primary")
+
+        with gr.Column():
+            gr.Markdown("### Output Visualizations")
+            hsv_histogram_display = gr.Image(label='HSV Histogram')
+            binary_mask_display = gr.Image(label="Segmentation Output")
+
+    update_button.click(
+        process_hsv_image, 
+        inputs=[image_input, lower, upper], 
+        outputs=[hsv_histogram_display, binary_mask_display]
+    )
+
+app.launch()

requirements.txt

@@ -0,0 +1,5 @@
+opencv-contrib-python
+opencv-python
+matplotlib
+pillow
+numpy

utils.py

@@ -0,0 +1,114 @@
+import cv2
+import numpy as np
+import os
+import matplotlib.pyplot as plt
+import io
+from PIL import Image
+import numpy as np
+
+def process_bgr_image(img, lb, ub, channel):
+
+    # Define the lower and upperbounds based on the image channel.
+    if channel == 'Blue':
+        lb = np.array([lb, 0, 0], np.uint8)
+        ub = np.array([ub, 255, 255], np.uint8)
+    elif channel == 'Green':
+        lb = np.array([0, lb, 0], np.uint8)
+        ub = np.array([255, ub, 255], np.uint8)
+    elif channel == 'Red':
+        lb = np.array([0, 0, lb], np.uint8)
+        ub = np.array([255, 255, ub], np.uint8)
+
+    # Calculate the histograms for the BGR color space and the Hue channel in the HSV image.
+    hist_b, hist_g, hist_r = calculateHistogram(img, mode='BGR')
+
+    # Plot the histograms
+    hist_bgr = plotHistogram((hist_b, hist_g, hist_r), mode='BGR')
+
+    # Calculate the mask for the given bounds from hsv image
+    op = calculateMask(img, lb, ub, mode='BGR')
+
+    return  hist_bgr, op
+
+def process_hsv_image(img, lb, ub):
+    
+    # Conver the image to the HSV color space
+    img_hsv = cv2.cvtColor(img, cv2.COLOR_BGR2HSV)
+
+    # Define the lower and upperbounds for Hue
+    lb = np.array([lb, 0, 0], np.uint8)
+    ub = np.array([ub, 255, 255], np.uint8)
+
+    # Calculate the histograms for the BGR color space and the Hue channel in the HSV image.
+    hist_hue = calculateHistogram(img_hsv, mode='HSV')
+
+    # Plot the histogram.
+    hist_hsv = plotHistogram(hist_hue, mode='HSV')
+
+    # Calculate the mask for the given bounds from hsv image
+    op = calculateMask((img_hsv, img), lb, ub, mode='HSV')
+
+    return hist_hsv, op
+
+
+def calculateHistogram(img, mode):
+
+    if mode == 'BGR':
+        hist_b = cv2.calcHist([img], [0], None, [256], [0,255])
+        hist_g = cv2.calcHist([img], [1], None, [256], [0,255])
+        hist_r = cv2.calcHist([img], [2], None, [256], [0,255])
+
+        return hist_b, hist_g, hist_r
+    
+    elif mode == 'HSV':
+
+        hist_hue = cv2.calcHist([img], [0], None, [180], [0,180])
+        return hist_hue
+
+def plotHistogram(hists, mode):
+
+    if mode == 'BGR':
+        hist_b, hist_g, hist_r = hists
+        plt.figure(figsize=(10,2))
+        plt.plot(hist_b, color='b', label='Blue')
+        plt.plot(hist_g, color='g', label='Green')
+        plt.plot(hist_r, color='r', label='Red')
+        plt.xlim([0, 255])
+        plt.legend()
+        buf_bgr= io.BytesIO()
+        plt.savefig(buf_bgr, format='png')
+        buf_bgr.seek(0)
+        hist_bgr = Image.open(buf_bgr)
+        plt.close()
+        return hist_bgr
+    
+    if mode == 'HSV':
+
+        hist_hue = hists
+        plt.figure(figsize=(10,2))
+        plt.plot(hist_hue, color='black', label='Black')
+        plt.xlim([0, 180])
+        buf_hsv = io.BytesIO()
+        plt.savefig(buf_hsv, format='png')
+        buf_hsv.seek(0)
+        hist_hsv = Image.open(buf_hsv)
+        plt.close()
+        return hist_hsv
+
+def calculateMask(imgs, lb, ub, mode):
+
+    op = None
+
+    if mode == 'BGR':
+        img = imgs
+        mask_1d = cv2.inRange(img, lb, ub)
+        mask_3d = cv2.merge([mask_1d, mask_1d, mask_1d])
+        op = cv2.bitwise_and(img, mask_3d)
+    
+    if mode == 'HSV':
+        img_hsv, img = imgs
+        mask_1d = cv2.inRange(img_hsv, lb, ub)
+        mask_3d = cv2.merge([mask_1d, mask_1d, mask_1d])
+        op = cv2.bitwise_and(img, mask_3d)
+
+    return op