import streamlit as st import cv2 import numpy as np from PIL import Image import pandas as pd import plotly.express as px import plotly.graph_objects as go from ultralytics import YOLO import time st.set_page_config( page_title="Blood Cell Detection App", page_icon="đŸ”Ŧ", layout="wide", initial_sidebar_state="expanded" ) st.markdown(""" """, unsafe_allow_html=True) st.markdown("

Blood Cell Detection System

", unsafe_allow_html=True) tab1, tab2, tab3 = st.tabs(["📊 Analysis", "ℹī¸ About", "❓ Help"]) with tab1: @st.cache_data def preprocess_image(uploaded_file): try: bytes_data = uploaded_file.getvalue() img = cv2.imdecode(np.frombuffer(bytes_data, np.uint8), cv2.IMREAD_COLOR) if img is None: raise ValueError("Invalid image file") return cv2.cvtColor(img, cv2.COLOR_BGR2RGB) except Exception as e: st.error(f"Error preprocessing image: {str(e)}") return None @st.cache_resource def load_model(): try: model = YOLO("bccd_yolov10_best.pt") return model except Exception as e: st.error(f"Error loading model: {str(e)}") return None def perform_inference(model, image): try: results = model.predict(image, conf=0.25) return results[0] except Exception as e: st.error(f"Error during inference: {str(e)}") return None def create_visualization(image, results): if not results.boxes: return None fig = go.Figure() fig.add_trace(go.Image(z=image)) boxes = results.boxes.xyxy.cpu().numpy() classes = results.boxes.cls.cpu().numpy() confs = results.boxes.conf.cpu().numpy() class_names = ["RBC", "WBC", "Platelets"] colors = ["#FF5252", "#2196F3", "#4CAF50"] for box, cls, conf in zip(boxes, classes, confs): x0, y0, x1, y1 = box cls_name = class_names[int(cls)] color = colors[int(cls)] fig.add_shape(type="rect", x0=x0, y0=y0, x1=x1, y1=y1, line=dict(color=color, width=2)) fig.add_annotation(x=x0, y=y0-5, text=f"{cls_name} ({conf:.2f})", showarrow=False, font=dict(color="white", size=10), bgcolor=color, bordercolor=color) fig.update_layout(width=700, height=500, margin=dict(l=0, r=0, b=0, t=0), showlegend=False) fig.update_xaxes(showticklabels=False, showgrid=False) fig.update_yaxes(showticklabels=False, showgrid=False) return fig def create_cell_counts_chart(results): if not results.boxes: return None classes = results.boxes.cls.cpu().numpy() class_names = ["RBC", "WBC", "Platelets"] counts = {name: sum(1 for cls in classes if class_names[int(cls)] == name) for name in class_names} df = pd.DataFrame({"Cell Type": list(counts.keys()), "Count": list(counts.values())}) fig = px.bar(df, x="Cell Type", y="Count", color="Cell Type", color_discrete_sequence=["#FF5252", "#2196F3", "#4CAF50"], text=df["Count"], height=400) fig.update_layout(title="Cell Count Distribution", xaxis_title="", yaxis_title="Count") fig.update_traces(textposition="outside") return fig # Main Content st.markdown("
", unsafe_allow_html=True) model = load_model() if not model: st.stop() st.markdown("

Upload Image

", unsafe_allow_html=True) uploaded_file = st.file_uploader("Choose an image (JPG, JPEG, PNG)", type=["jpg", "jpeg", "png"]) st.markdown("
", unsafe_allow_html=True) if uploaded_file: st.markdown("
", unsafe_allow_html=True) col1, col2 = st.columns(2) with col1: image = preprocess_image(uploaded_file) if image is None: st.stop() st.image(image, caption="Uploaded Image", use_container_width=True) with col2: if st.button("🔍 Detect Blood Cells"): with st.spinner("Analyzing..."): time.sleep(1) # Simulate processing results = perform_inference(model, image) if results and results.boxes: vis_fig = create_visualization(image, results) st.plotly_chart(vis_fig, use_container_width=True) # Metrics classes = results.boxes.cls.cpu().numpy() class_names = ["RBC", "WBC", "Platelets"] counts = {name: sum(1 for cls in classes if class_names[int(cls)] == name) for name in class_names} cols = st.columns(4) cols[0].markdown(f"
{len(classes)}
Total Cells
", unsafe_allow_html=True) cols[1].markdown(f"
{counts['RBC']}
RBC
", unsafe_allow_html=True) cols[2].markdown(f"
{counts['WBC']}
WBC
", unsafe_allow_html=True) cols[3].markdown(f"
{counts['Platelets']}
Platelets
", unsafe_allow_html=True) count_fig = create_cell_counts_chart(results) st.plotly_chart(count_fig, use_container_width=True) st.markdown("

Detailed Results

", unsafe_allow_html=True) data = [{"ID": i+1, "Cell Type": f"{class_names[int(cls)]}", "Confidence": f"{conf*100:.2f}%"} for i, (cls, conf) in enumerate(zip(classes, results.boxes.conf.cpu().numpy()))] st.write(pd.DataFrame(data).to_html(escape=False, index=False), unsafe_allow_html=True) st.download_button("📊 Export as CSV", pd.DataFrame({"Cell Type": [class_names[int(cls)] for cls in classes], "Confidence": [f"{conf*100:.2f}%" for conf in results.boxes.conf.cpu().numpy()]}).to_csv(index=False), "results.csv", "text/csv") else: st.warning("No cells detected or analysis failed.") st.markdown("
", unsafe_allow_html=True) with tab2: st.markdown("
", unsafe_allow_html=True) st.markdown("

About

", unsafe_allow_html=True) st.markdown(""" This app uses YOLOv10 to detect Red Blood Cells (RBC), White Blood Cells (WBC), and Platelets in blood smear images. Upload an image in the Analysis tab to see counts, visualizations, and detailed results. """) st.markdown("
", unsafe_allow_html=True) with tab3: st.markdown("
", unsafe_allow_html=True) st.markdown("

Help

", unsafe_allow_html=True) st.markdown(""" - **Upload**: Select a JPG, JPEG, or PNG image in the Analysis tab. - **Detect**: Click "Detect Blood Cells" to analyze. - **Results**: View counts and export data as CSV. - **Support**: Email ss93134041@gmail.com for issues. """) st.markdown("
", unsafe_allow_html=True)