import cv2 import base64 from dotenv import load_dotenv load_dotenv() def capture_image() -> str: """ Captures one frame from the default webcam, resizes it, encodes it as Base64 JPEG (raw string) and returns it. """ for idx in range(4): cap = cv2.VideoCapture(idx, cv2.CAP_ANY) if cap.isOpened(): for _ in range(10): # Warm up cap.read() ret, frame = cap.read() cap.release() if not ret: continue cv2.imwrite("sample.jpg", frame) # Optional ret, buf = cv2.imencode('.jpg', frame) if ret: return base64.b64encode(buf).decode('utf-8') raise RuntimeError("Could not open any webcam (tried indices 0-3)") #capture_image() from groq import Groq def analyze_image_with_query(query: str) -> str: """ Expects a string with 'query'. Captures the image and sends the query and the image to to Groq's vision chat API and returns the analysis. """ img_b64 = capture_image() model="meta-llama/llama-4-maverick-17b-128e-instruct" if not query or not img_b64: return "Error: both 'query' and 'image' fields required." client=Groq() messages=[ { "role": "user", "content": [ { "type": "text", "text": query }, { "type": "image_url", "image_url": { "url": f"data:image/jpeg;base64,{img_b64}", }, }, ], }] chat_completion=client.chat.completions.create( messages=messages, model=model ) return chat_completion.choices[0].message.content #query = "How many people do you see?" #print(analyze_image_with_query(query))