@extends('layouts.app')

@section('title', 'Deteksi Real-Time')

@section('contents')
<section class="about section" id="about">
    <h2 class="section-title text-center">Deteksi Pose Jurus 1 Tangan Kosong IPSI</h2>
    <div class="about__container" style="display: flex; flex-wrap: wrap; justify-content: center; gap: 20px;">
        <div class="video-container" style="position: relative;">
            <video id="webcam" width="640" height="480" autoplay playsinline muted></video>
            <canvas id="output-canvas" width="640" height="480"
                    style="position: absolute; top: 0; left: 0;"></canvas>
        </div>
        <div id="prediction-result" class="prediction-box">
            <p>Prediksi: <span id="pose-label">-</span></p>
            <p>Probability: <span id="pose-accuracy">-</span></p>
        </div>
    </div>
</section>
@endsection

@section('script')
<script src="https://code.jquery.com/jquery-3.6.0.min.js" defer></script>
<script src="https://cdn.jsdelivr.net/npm/@tensorflow/tfjs"></script>
<script src="https://cdn.jsdelivr.net/npm/@mediapipe/pose"></script>
<script src="https://cdn.jsdelivr.net/npm/@mediapipe/camera_utils"></script>

<script>
const poseClasses = [
    'A1_benar', 'A1_salah','A2_benar', 'A2_salah', 'A3_benar', 'A3_salah',
    'A4_benar', 'A4_salah', 'A5_benar', 'A5_salah', 'A6_benar', 'A6_salah',
    'A7_benar', 'A7_salah'
];

const videoElement = document.getElementById('webcam');
const canvasElement = document.getElementById('output-canvas');
const ctx = canvasElement.getContext('2d');
const poseLabelElement = document.getElementById('pose-label');
const poseAccuracyElement = document.getElementById('pose-accuracy');

let model = null;
let poseDetector = null;
let camera = null;

async function loadModel() {
    try {
        model = await tf.loadLayersModel('/models/model.json');
        console.log('✅ Model loaded');
    } catch (err) {
        console.error('❌ Gagal load model:', err);
        alert('Model gagal dimuat');
    }
}

function setupPose() {
    poseDetector = new Pose({
        locateFile: (file) => `https://cdn.jsdelivr.net/npm/@mediapipe/pose/${file}`
    });

    poseDetector.setOptions({
        modelComplexity: 1,
        smoothLandmarks: true,
        enableSegmentation: false,
        minDetectionConfidence: 0.5,
        minTrackingConfidence: 0.5
    });

    poseDetector.onResults(onPoseResults);
}

function onPoseResults(results) {
    if (!results.poseLandmarks) return;

    ctx.clearRect(0, 0, canvasElement.width, canvasElement.height);
    ctx.drawImage(results.image, 0, 0, canvasElement.width, canvasElement.height);

    drawLandmarks(results.poseLandmarks);

    const landmarks = results.poseLandmarks.map(landmark => [
        landmark.x * canvasElement.height,
        landmark.y * canvasElement.height,
        landmark.z * canvasElement.width
    ]).flat();

    predictPose(landmarks);
}

function drawLandmarks(landmarks) {
    drawConnections(landmarks);
    ctx.fillStyle = '#FF0000';
    landmarks.forEach(landmark => {
        ctx.beginPath();
        ctx.arc(landmark.x * canvasElement.width, landmark.y * canvasElement.height, 4, 0, 2 * Math.PI);
        ctx.fill();
    });
}

function drawConnections(landmarks) {
    const connections = [
        [0, 1], [1, 2], [2, 3], [3, 7], [0, 4], [4, 5], [5, 6], [6, 8],
        [11, 13], [13, 15], [15, 17], [15, 19], [15, 21],
        [12, 14], [14, 16], [16, 18], [16, 20], [16, 22],
        [11, 12], [11, 23], [12, 24], [23, 24],
        [23, 25], [25, 27], [27, 29], [27, 31],
        [24, 26], [26, 28], [28, 30], [28, 32]
    ];

    ctx.strokeStyle = '#00FF00';
    ctx.lineWidth = 2;

    connections.forEach(([i, j]) => {
        if (landmarks[i] && landmarks[j]) {
            ctx.beginPath();
            ctx.moveTo(landmarks[i].x * canvasElement.width, landmarks[i].y * canvasElement.height);
            ctx.lineTo(landmarks[j].x * canvasElement.width, landmarks[j].y * canvasElement.height);
            ctx.stroke();
        }
    });
}

async function predictPose(landmarks) {
    if (!model) return;

    const inputTensor = tf.tensor2d([landmarks]);
    const prediction = model.predict(inputTensor);
    const [predictedClass] = await prediction.argMax(1).data();
    const confidence = await prediction.max(1).data();

    poseLabelElement.textContent = poseClasses[predictedClass] || 'Unknown';
    poseAccuracyElement.textContent = `${(confidence[0]).toFixed(4)}`;

    inputTensor.dispose();
    prediction.dispose();
}

function startCamera() {
    camera = new Camera(videoElement, {
        onFrame: async () => {
            await poseDetector.send({ image: videoElement });
        },
        width: 640,
        height: 480
    });
    camera.start();
}

async function init() {
    await loadModel();
    setupPose();
    startCamera();
}

document.addEventListener('DOMContentLoaded', init);
</script>

<style>
.section-title {
    text-align: center;
    font-size: 1.8rem;
    margin-bottom: 20px;
    color: #001f3f; /* biru dongker */
}

.about__container {
    display: flex;
    flex-wrap: wrap;
    justify-content: center;
    gap: 20px;
}

.video-container {
    position: relative;
    width: 640px;
    height: 480px;
    border: 2px solid #ccc;
    border-radius: 10px;
    overflow: hidden;
}

.prediction-box {
    flex-shrink: 0;
    width: 280px;
    height: fit-content;
    padding: 20px;
    background: #f4f4f4;
    text-align: center;
    border-radius: 10px;
    font-size: 1.2rem;
    box-shadow: 0 2px 6px rgba(0,0,0,0.1);
}

#pose-label {
    font-weight: bold;
    color: #2c3e50;
}

#pose-accuracy {
    font-weight: bold;
    color: #27ae60;
}
</style>
@endsection