Update pages/Roadmap of ML Project.py

pages/Roadmap of ML Project.py

@@ -495,141 +495,6 @@ elif st.session_state.step == "data_collection":
                         notebook_html = f.read()
                     st.components.v1.html(notebook_html, height=500, scrolling=True)
 
-                image_augumentation_button= st.button('Image Augumentation')
-                if image_augumentation_button:
-                    # Streamlit header
-                    st.markdown('<h1><center>Affine Transformatiosn on Image</center></h1>', unsafe_allow_html=True)
-                    st.write("""Affine transformations are used in image augmentation to manipulate and alter images by applying operations such as translation, rotation, scaling, and shearing. 
-                    These transformations help to generate new images from the existing ones, thereby increasing the dataset diversity.""")
-                    # Image augmentation functions
-                    def rotate_image(image, angle):
-                        (h, w) = image.shape[:2]  # Get image dimensions
-                        center = (w // 2, h // 2)  # Define the center of the image
-                        M = cv2.getRotationMatrix2D(center, angle, 1.0)  # Rotation matrix
-                        rotated = cv2.warpAffine(image, M, (w, h),borderMode=cv2.BORDER_REFLECT)  # Apply rotation
-                        return rotated
-                    
-                    def shifting(image, tx, ty):
-                        t_matrix = np.array([[1, 0, tx],
-                                              [0, 1, ty]], dtype=np.float32)
-                        t_img = cv2.warpAffine(image, t_matrix, (image.shape[1], image.shape[0]))  # Adjust size to original
-                        return t_img
-                    
-                    def scaling(image,fx,fy):
-                        scaled_image=cv2.resize(image,None,fx=fx, fy=fy)
-                        return scaled_image
-                    
-                    def shearing(image,shx,shy,tsx,tsy):
-                        shearing_matrix = np.array([[1,shx,tsx],[shy,1,tsy]],dtype= np.float32)
-                        sheared_image=cv2.warpAffine(image,shearing_matrix,(image.shape[1], image.shape[0]))
-                        return sheared_image
-                    
-                    def cropping(image,x1,x2,y1,y2):
-                        cropped_image=image[x1:x2,y1:y2]
-                        return cropped_image
-                        
-                        
-                    
-                    # Streamlit app
-                    st.markdown("<b>Upload an image and apply various augmentation techniques.</b>", unsafe_allow_html=True)
-                    
-                    # File uploader for images
-                    uploaded_file = st.file_uploader("Choose an image...", type=["jpg", "png", "jpeg"])
-                    
-                    if uploaded_file is not None:
-                        # Read the image using OpenCV directly (ensure it's in BGR format)
-                        image_bgr = np.asarray(bytearray(uploaded_file.read()), dtype=np.uint8)
-                        image_bgr = cv2.imdecode(image_bgr, cv2.IMREAD_COLOR)  # Read as BGR image
-                    
-                        # Display the uploaded image (converted to RGB for Streamlit)
-                        image_rgb = cv2.cvtColor(image_bgr, cv2.COLOR_BGR2RGB)  # Convert to RGB for display
-                        st.image(image_rgb, caption="Uploaded Image", use_container_width=True)
-                    
-                        st.subheader('Rotation')
-                    
-                        # Rotation slider
-                        rotate = st.slider("Rotate Angle", 0, 180, 45)
-                    
-                        # Apply rotation based on the slider value
-                        if rotate != 0:
-                            # Apply rotation
-                            rotated_image_bgr = rotate_image(image_bgr, rotate)
-                    
-                            # Convert back to RGB for Streamlit
-                            rotated_image_rgb = cv2.cvtColor(rotated_image_bgr, cv2.COLOR_BGR2RGB)
-                            
-                            # Display the rotated image
-                            st.image(rotated_image_rgb, caption=f"Rotated Image ({rotate}°)", use_container_width=True)
-                        else:
-                            st.write("No rotation applied.")
-                    
-                        st.subheader('Translation')
-                        # Shifting sliders
-                        Tx = st.slider('tx distance', -100, 100, 0)  # Allow both positive and negative values
-                        Ty = st.slider('ty distance', -100, 100, 0)
-                    
-                        # Apply shifting based on the sliders
-                        if Tx != 0 or Ty != 0:
-                            shifted_image = shifting(image_bgr, Tx, Ty)
-                            shifted_image_rgb = cv2.cvtColor(shifted_image, cv2.COLOR_BGR2RGB)
-                            st.image(shifted_image_rgb, caption=f"Translated Image (Tx: {Tx}, Ty: {Ty})", use_container_width=True)
-                        else:
-                            st.write("No Translation applied.")
-                    
-                        st.subheader('Scaling')
-                        
-                        #scaling
-                        fx=st.slider('fx distance', 0.5,1.5, 0.0,)
-                        fy=st.slider('fy distance', 0.5, 1.5, 0.0)
-                        # apply scaling based on sliders
-                        if fx!=0 or fy!=0:
-                            scaling_image = scaling(image_bgr,fx,fy)
-                            scaling_image_rgb=cv2.cvtColor(scaling_image, cv2.COLOR_BGR2RGB)
-                            st.image(scaling_image_rgb, caption=f'Scaled Image (fx:{fx},fy:{fy})',use_container_width=True)
-                        else:
-                            st.write('No scaling Applied')
-                    
-                    
-                        st.subheader('Shearing')
-                        
-                        # Shearing sliders
-                        shx = st.slider('Shearing on x-axis', 0.0, 3.0, 0.0, key='slider_shx')
-                        shy = st.slider('Shearing on y-axis', 0.0, 3.0, 0.0, key='slider_shy')
-                        tsx = st.slider('Translation on x-axis (tx)', -100, 100, 0, key='slider_tsx')
-                        tsy = st.slider('Translation on y-axis (ty)', -100, 100, 0, key='slider_tsy')
-                        
-                        # Apply shearing on the image
-                        if shx != 0 or shy != 0:
-                            # Assuming 'shearing' is a function that applies the shear transformation
-                            shearing_image = shearing(image_bgr, shx, shy, tsx, tsy)
-                            
-                            # Convert the sheared image to RGB for display in Streamlit
-                            shearing_image_rgb = cv2.cvtColor(shearing_image, cv2.COLOR_BGR2RGB)
-                            
-                            # Display the sheared image
-                            st.image(
-                                shearing_image_rgb,
-                                caption=f'Sheared image (sx: {shx}, sy: {shy})',
-                                use_container_width=True
-                            )
-                        else:
-                            st.write("No Shearing applied.")
-                    
-                        st.subheader('Cropping')
-                        
-                        # Cropping sliders
-                        x1 = st.slider('x1 value', 0, 1000, 0, key='slider_x1')
-                        x2 = st.slider('x2 value', 0, 1000, 0, key='slider_x2')
-                        y1 = st.slider('y1 value', 0, 1000, 0, key='slider_y1')
-                        y2 = st.slider('y2 value', 0, 1000, 0, key='slider_y2')
-                        
-                        if x1 != 0 or x2 != 0 or y1 != 0 or y2 != 0:
-                            cropping_image = cropping(image_bgr, x1, x2, y1, y2)
-                            cropping_img_rgb = cv2.cvtColor(cropping_image, cv2.COLOR_BGR2RGB)
-                            # Display the cropped image with proper caption formatting
-                            st.image(cropping_img_rgb, caption=f'cropped image, image[{x1}:{x2}, {y1}:{y2}]', use_container_width=True)
-                        else:
-                            st.write('No Cropping Applied')
                 
                     
             Video_button= st.button("Videos")