3d-camera-interaction
Provides corrected math and logic for 3D model manipulation in Three.js.
Install
mkdir -p .claude/skills/3d-camera-interaction && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/1819" && unzip -o skill.zip -d .claude/skills/3d-camera-interaction && rm skill.zipInstalls to .claude/skills/3d-camera-interaction
Activation
This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.
Three.js 中处理 3D 模型拖拽、缩放、边界检测的正确方法。解决鼠标移动与模型移动不同步、缩放后只能看到模型一部分等问题。Key capabilities
- →Calculate pixel-to-world conversion based on camera distance
- →Implement correct zoom-to-center logic
- →Apply boundary constraints using screen-space projection
- →Synchronize mouse movement with 3D model translation
How it works
It dynamically maps screen-space pixel movement to world-space coordinates by calculating the viewport height at the current camera distance.
Inputs & outputs
When to use 3d-camera-interaction
- →Fix 3D model drag jitter
- →Implement correct zoom-to-center logic
- →Apply 3D boundary constraints
About this skill
3D 相机交互:拖拽与边界检测
症状
- 缩放后拖拽模型,鼠标移动 100px 但模型移动的屏幕距离不是 100px
- 放大模型后只能看到腿/身体的一部分,无法正常平移
- 拖动开始时模型位置"跳变"
根本原因
原因 1: 固定 panSpeed 导致移动不同步
问题: 使用固定的 panSpeed = 0.01 进行平移计算
// ❌ 错误方式
const panSpeed = 0.01;
newPosition.add(right.multiplyScalar(deltaX * panSpeed));
为什么发生: 相机距离变化时,同样的世界空间距离在屏幕上的像素表现不同。距离近时像素多,距离远时像素少。
解决方案: 根据相机距离和 FOV 动态计算像素→世界空间的映射
// ✅ 正确方式:动态计算
const cameraDistance = camera.position.distanceTo(modelCenter);
const fov = camera.fov * (Math.PI / 180);
const screenHeight = renderer.domElement.clientHeight;
const screenWidth = renderer.domElement.clientWidth;
// 在相机距离处,视口的世界空间高度
const worldHeight = 2 * Math.tan(fov / 2) * cameraDistance;
const worldWidth = worldHeight * (screenWidth / screenHeight);
// 每像素对应的世界空间距离
const pixelToWorldX = worldWidth / screenWidth;
const pixelToWorldY = worldHeight / screenHeight;
// 应用:鼠标移动的像素 × 每像素对应的世界空间距离
newPosition.add(right.multiplyScalar(deltaX * pixelToWorldX));
newPosition.add(up.multiplyScalar(-deltaY * pixelToWorldY));
原因 2: 基于中心点的边界限制
问题: 使用模型中心点的 NDC 坐标判断是否出界
// ❌ 错误方式:限制中心点位置
const ndc = position.clone().project(camera);
if (ndc.y > 0.2) clampedY = 0.2; // 限制顶部
为什么发生: 模型放大后,中心点在屏幕中心,但身体大部分已超出屏幕。限制中心点 = 限制只能看到身体中间部分。
解决方案: 计算模型在屏幕上的可见区域(像素),只在可见区域过小时才校正
// ✅ 正确方式:基于可见像素
const MIN_VISIBLE_PIXELS = 50;
// 1. 计算模型包围盒并投影到屏幕
const box = new THREE.Box3().setFromObject(vrm.scene);
const corners = [/* 8个顶点 */];
let modelMinX = Infinity, modelMaxX = -Infinity;
let modelMinY = Infinity, modelMaxY = -Infinity;
corners.forEach(corner => {
const projected = corner.clone().project(camera);
const screenX = (projected.x * 0.5 + 0.5) * screenWidth;
const screenY = (-projected.y * 0.5 + 0.5) * screenHeight;
// 更新边界...
});
// 2. 计算可见区域
const visibleWidth = Math.max(0, Math.min(screenWidth, modelMaxX) - Math.max(0, modelMinX));
const visibleHeight = Math.max(0, Math.min(screenHeight, modelMaxY) - Math.max(0, modelMinY));
const visiblePixels = visibleWidth * visibleHeight;
// 3. 只在可见区域太小时校正
if (visiblePixels < MIN_VISIBLE_PIXELS) {
// 将模型拉回可见区域
}
关键公式
像素到世界空间转换
worldHeight = 2 × tan(fov/2) × cameraDistance
pixelToWorld = worldHeight / screenHeight
世界坐标到屏幕坐标
const ndc = worldPos.clone().project(camera);
const screenX = (ndc.x * 0.5 + 0.5) * screenWidth;
const screenY = (-ndc.y * 0.5 + 0.5) * screenHeight; // Y 轴反向
关键经验
- 📐 相机距离影响一切: 所有像素↔世界空间的转换都需要考虑相机距离
- 🔲 使用包围盒而非中心点: 边界检测应基于模型实际占用的屏幕区域
- 🔄 与 2D 保持一致: Live2D/VRM 等不同类型模型应使用相同的交互逻辑阈值
When not to use it
- →When using fixed pan speeds for camera interaction
Prerequisites
Limitations
- →Requires accurate camera distance calculation
- →Boundary detection depends on model bounding box
How it compares
It solves the jitter and synchronization issues inherent in fixed-speed panning by accounting for camera FOV and distance.
Compared to similar skills
3d-camera-interaction side by side with the closest alternatives in the catalog.
| Skill | Installs | Updated | Safety | Difficulty |
|---|---|---|---|---|
| 3d-camera-interaction (this skill) | 3 | 6mo | No flags | Advanced |
| vue-debug-guides | 4 | 6mo | No flags | Intermediate |
| agentation | 6 | 5mo | Review | Beginner |
| vrm-springbone-physics | 0 | 6mo | No flags | Intermediate |
Try saying
Example prompts that trigger this skill in your AI assistant.
More by Project-N-E-K-O
View all by Project-N-E-K-O →You might also like
vue-debug-guides
vuejs-ai
Vue 3 debugging and error handling for runtime errors, warnings, async failures, and SSR/hydration issues. Use when diagnosing or fixing Vue issues.
agentation
benjitaylor
Add Agentation visual feedback toolbar to a Next.js project
vrm-springbone-physics
Project-N-E-K-O
Debugging and fixing VRM SpringBone physics issues in three-vrm, including hair/clothing physics that flies upward, sticks out horizontally, or behaves unnaturally.
debugging-skill
bitovi
Always start a fresh browser session after any file change, walk through the full user flow, and monitor for errors before proceeding with further work.
expect
zakad00e2
Use when editing .tsx/.jsx/.css/.html, React components, pages, routes, forms, styles, or layouts. Also when asked to test, verify, validate, QA, find bugs, check for issues, or fix expect-cli failures.
frontend-api-integration-patterns
TJSNDHU
Production-ready patterns for integrating frontend applications with backend APIs, including race condition handling, request cancellation, retry strategies, error normalization, and UI state management.