AI短视频赚钱:爆款视频拆解之图解万物

大家好,欢迎来到烁变之路!
最近刷视频看到一个视频题材挺不错,图解万物,以海报的形式科普一些动物宠物。没有特别火爆,但胜在流量长久,很稳定。
我们就去拆解一下这种视频是怎么制作的。

一、提示词制作

这是这个视频制作的关键点,没有好的提示词,是做不出来那种海报效果的,我用的是谷歌。有条件的最后用这个,不行也可以用国内的豆包,它们来代替,只不过出图质量会比较差,要多调整几次。
我们先做一个出提示词的智能体,点击“gem”豆包里就是智能体一样的。
(源于小太子奶)输入提示词:
角色设定: 你是一位精通 19 世纪自然科学、博物学插画以及 AI 图像工程的提示词专家。你的任务是根据用户提供的“宠物名称”,自动生成一段结构极其复杂、细节极其丰富的 Midjourney/DALL-E3 提示词,风格严格统一为“复古自然历史百科全书信息图”
提示词生成逻辑:
核心视觉: 将宠物置于中心,强调写实毛发、肌肉感及灵动的神态环境重构:自动匹配一个符合该宠物品种特征的中式复古室内或庭院背景。模块化布局: 严格按照以下三个板块填充专业科普内容:包含体型示意、被毛图解、具体参数(体长/体重/寿命)、性格标签左侧:右侧:包含该宠物的行为场景(高频/互动/探索)、人宠关系、以及3个幽默的趣味标
底部:包含绿色的“优点”和红色的“缺点”对比模块。质感渲染:强制使用“泛黄旧报纸、边缘磨损撕裂、手绘水墨色彩、暖调复古印刷等关键词。
输出格式: 只输出一段完整的、高质量的英文 Prompt,并在 Prompt 结束后,用中文简要概括你为这个品种设计的“趣味标签”是什么。触发指令: 当用户输入任意宠物名字时,立即开始生成。
输入要制作的动物名称,然后发送(不能用快速,要用思考,快速出来的提示词比较简单)
复制生成的提示词:
柴犬:
A complex 19th-century natural history encyclopedia infographic dedicated to the "Shiba Inu" (Canis lupus familiaris). Centerpiece: A highly realistic Shiba Inu with dense, textured orange-red fur and iconic "Urajiro" white markings, showcasing muscular agility and a spirited, squinting expression, positioned within a vintage Chinese-style courtyard featuring weathered red lacquered pillars and a stone-carved lion statue. Left Module: Technical anatomical diagrams including a body scale silhouette, a microscopic cross-section illustration of the double-layered coat, and a vintage data box labeled "Morphology" with parameters: "Length: 35-43cm", "Weight: 8-11kg", "Lifespan: 12-15 years", and temperament tags "Fastidious, Independent, Bold". Right Module: A series of small behavioral sketches illustrating "The Airplane Ear Greeting", "Human-Pet Interaction in a Tea House", and "The Stubborn Refusal to Walk". Features three decorative humorous labels: "The Grumpy Loaf", "Drama King of Canines", and "Sentient Marshmallow". Bottom Module: A horizontal comparative table titled "Ethical Evaluation" with a green "Pros" section listing "Innate Cleanliness" and "Silent Loyalty", and a red "Cons" section listing "Extreme Stubbornness" and "Seasonal Fur Blizzard". Aesthetic: Hand-drawn ink and watercolor style, warm vintage printing, printed on heavily yellowed, distressed old parchment with foxing, tea stains, and torn deckled edges, intricate scientific typography, 8k resolution, museum-quality lithograph. --ar 3:4 --v 6.0
阿拉斯加犬:
A complex 19th-century natural history encyclopedia infographic dedicated to the "Alaskan Malamute" (Canis lupus familiaris borealis). Centerpiece: A highly realistic, immensely powerfully built Alaskan Malamute with a thick, dense grey and white double coat and a plumed tail curled over its back, showcasing robust musculature and a dignified, friendly expression. It is positioned within a vintage arctic explorer's wooden cabin interior, filled with antique snowshoes, leather sled harnesses hanging on rough-hewn log walls, and a stone fireplace, looking out a frosted window at a snowy mountain landscape. Left Module: Technical anatomical diagrams including a comparative body scale silhouette next to a human explorer figure, a detailed cross-section illustration of the thermal double-layered pelage, and a vintage data box labeled "Morphology" with parameters: "Height: 58-71cm", "Weight: 34-50kg", "Lifespan: 10-14 years", and temperament tags "Affectionate, Loyal, Dignified, Playful". Right Module: A series of small behavioral sketches illustrating "Sled Pulling in Team Formation", "Human-Pet Interaction: The 'Woo-Woo' Howl Communication", and "Digging a Snow Den". Features three decorative humorous labels in ornate banners: "The Woolly Woo-Woo Machine", "Part-Time Sled Engine", and "Stubborn Snow Dozer". Bottom Module: A horizontal comparative table titled "Ethical Evaluation" with a green "Pros" section listing "Incredible Endurance" and "Friendly to Strangers", and a red "Cons" section listing "Heavy Seasonal Shedding" and "Strong Prey Drive & Pulling Instinct". Aesthetic: Hand-drawn ink and watercolor style, warm vintage printing, printed on heavily yellowed, distressed old parchment with foxing, tea stains, and torn deckled edges, intricate scientific typography, 8k resolution, museum-quality lithograph. --ar 3:4 --v 6.0
金毛 :
A complex 19th-century natural history encyclopedia infographic dedicated to the "Golden Retriever" (Canis lupus familiaris). Centerpiece: A highly realistic, robustly built Golden Retriever with a lustrous, dense, wavy golden-hued coat, showcasing active musculature and a gentle, intelligent, eager-to-please expression. It is positioned within a refined vintage Chinese-style scholar's study, surrounded by antique rosewood furniture, hanging calligraphy scrolls, and a fireplace, with a leather retrieving dummy rests near its paws. Left Module: Technical anatomical diagrams including a body scale silhouette comparison next to a human figure, a microscopic cross-section illustration of the water-repellent double pelage, and a vintage data box labeled "Morphology & Attributes" with parameters: "Height: 51-61cm", "Weight: 25-34kg", "Lifespan: 10-12 years", and temperament tags "Intelligent, Friendly, Devoted, Eager". Right Module: A series of small behavioral sketches illustrating "Retrieving Waterfowl from a Lake", "Human-Pet Interaction: The Gentle 'Soft Mouth' Holding an Egg", and "Excited Greeting Behavior". Features three decorative humorous labels in ornate banners: "The Aggressively Happy Greeter", "Chief Ball Executive Officer", and "Golden Fluff Generator". Bottom Module: A horizontal comparative table titled "Ethical Evaluation" with a green "Pros" section listing "Highly Trainable" and "Ideal Family Companion", and a red "Cons" section listing "Heavy Seasonal Shedding" and "High Exercise Requirements". Aesthetic: Hand-drawn ink and watercolor style, warm vintage printing, printed on heavily yellowed, distressed old parchment with foxing, tea stains, and torn deckled edges, intricate scientific typography, 8k resolution, museum-quality lithograph. --ar 3:4 --v 6.0
上面的智能体只能制作动物类的提示词,想要做别的内容的,可以自己修改智能体提示词。

二、图片制作:

打开作图AI,我继续用香蕉,没有账号的可以用免费的,网址:https://arena.ai/
输入提示词,点击生成
图片中有不满意的地方可以把提示词翻译成中文,进行修改。
我就做了三张,一般一个视频要有七八张左右的内容。

三、视频剪辑

打开剪映,拉入我们的我们做的图片。
添加上背景音乐,直接导出就可以了。
AI短视频赚钱:爆款视频拆解之图解万物
一个简单的作品就做出来了
今天的视频拆解就到这里。
AI短视频赚钱:爆款视频拆解之图解万物

声明:本文转载自 烁变之路,作者为 烁变之路,原文网址:https://shuobian.com