街景摄影高级撕纸艺术海报

Please turn each of the travel street view photos I upload into an independent advanced torn paper art editorial poster, not a multi-image collage, with each photo output individually. The overall composition adopts a 3:4 vertical dual-panel layout, with the top part retaining a clearly visible authentic original image, and the bottom part reconstructing the same street alley as a "torn paper alley reconstruction"; the height of the top and bottom areas is nearly 1:1, using a clear, horizontal boundary, without blurring or mutual penetration. Only output the front of the poster, do not generate photo frames, walls, or handheld mockups.

The upper part accounts for about 47%–50% of the image, faithfully retaining the original photo. Accurately maintain the street's observation angle, vanishing direction, road slope, buildings on both sides, eaves, doors and windows, signs, vehicles, plants, wires, the number and posture of people, as well as the original natural light and color atmosphere. Only apply slight travel magazine-style color grading and extremely thin film grain processing, it must still obviously be real photography. To adapt to the frame, the sky, road surface, or quiet environment can be naturally extended, but buildings must not be redrawn, people moved, vehicles changed, stretched, mirrored, or subjects copied.

For the lower part, do not apply a paper filter to the original image with the same framing, nor use rectangular photo fragment collages. First, select one main street vanishing line from the photo, the main facades on both sides, and one to three objects that best describe the location, such as pedestrians, vehicles, lanterns, umbrellas, wires, plants, signs, or roofs. Proactively delete repetitive window frames, complex crowds, irrelevant small items, and photographic noise, then reorganize these elements into a smaller, entity torn paper scene with clear outlines.

The redrawn scene occupies about 60%–82% of the lower area, can be centered or slightly to the side, with about 20%–35% continuous ivory-colored paper surface preserved around it. Readjust the size and visual hierarchy of the main subject according to the original photo, making the most recognizable buildings, objects, or figures the paper anchors; do not copy the complete composition of the photo. A small torn paper circular window can be added to the white space to enlarge a real detail from the original photo, connected to the main scene with a graphite arrow.

Use hand-torn matte fiber paper, thin washi paper, corrugated paper, kraft paper, and a small amount of old printed paper to build facades, road surfaces, eaves, plants, skies, and characters piece by piece. All main shapes must have clearly visible fiber-torn edges, slight overlaps, paper thickness, curled corners, and short, natural physical shadows. Paper edges can be irregular, but the alley's outer contour, road direction, building boundaries, and subject identity must be clear. Forbidden are smooth vector slices, laser-cut edges, and flat digital shadows.

The color palette is completely extracted from the photo above, summarizing complex colors into five to seven grayish paper colors, with the real main colors of walls, roads, roofs, plants, and characters as the skeleton, retaining only one instance of rusty red, cobalt blue, olive green, or other characteristic colors from the original photo as an emphasis. Do not apply fixed Japanese travel color palettes to all locations.

Add a set of pencil inscriptions in the continuous bare paper area. Based on the locatable location, road structure, light, weather, or action in the photo, extract a title of 1–3 English words; add the field "PAPER STUDY 01", with multiple photos numbered sequentially; then add a row of 5–12 English words of observation sentences. The title should be clearly readable in normal chat previews, while fields and observation sentences are smaller but cannot be reduced to gray lines. The text is written directly on the paper like fine graphite or charcoal, can be accompanied by an arrow, and does not use smooth digital fonts. When the location cannot be confirmed, use the scene theme; do not fabricate place names, dates, or coordinates.

The overall piece should look like a page of paper-based work between architectural observation, material experimentation, and travel publications: the alley is identifiable at a glance, the paper layers are real, boundaries are clear, information is reduced, and there is enough white space to breathe.

Avoid scrapbooking nine-grid layouts, rectangular photo collages, isomorphic paper filters, full-background coverage, children's clip art, generic travel icons, fictional landmarks, author names, account names, logos, QR codes, watermarks, fake Chinese, and gibberish.

请将我上传的每一张旅行街景照片分别制作成一张独立的高级撕纸艺术编辑海报，不多图拼接，每张照片单独输出。整体采用3:4竖版双联构图，上方保留清楚可见的真实原图，下方将同一条街巷重构为“撕纸街巷重构”；上下区域高度接近1:1，使用清晰、水平的边界，不雾化、不互相渗透。只输出海报正面，不生成相框、墙面或手持Mockup。

上半部分约占画面47%–50%，忠实保留原始照片。准确保持街巷的观察角度、消失方向、道路坡度、两侧建筑、屋檐、门窗、招牌、车辆、植物、电线、人物数量与姿态，以及原有的自然光影和色彩氛围。只进行轻微的旅行杂志式调色和极薄胶片颗粒处理，仍必须明显是真实摄影。为适配画幅，可自然延展天空、路面或安静环境，但不得重画建筑、移动人物、改变车辆、拉伸、镜像或复制主体。

下半部分不要把原图以相同取景套上纸张滤镜，也不要使用矩形照片碎片拼贴。先从照片中选择一条最主要的街巷消失线、两侧主要立面，以及一至三项最能说明地点的物件，例如行人、车辆、灯笼、雨伞、电线、植物、招牌或屋顶。主动删除重复窗格、复杂人群、无关小物和摄影噪声，再将这些元素重新组织为一个缩小、具有清楚轮廓的实体撕纸场景。

转绘场景占下区约60%–82%，可居中或略微偏侧，四周保留约20%–35%连续象牙色纸面。根据原图重新调整主体大小和视觉层级，让最有辨识度的建筑、物件或人物成为纸面锚点，不要照搬照片的完整构图。可在留白中加入一个小型撕纸圆窗，放大一处来自原图的真实细节，并用一根石墨箭头与主场景连接。

使用手工撕开的哑光纤维纸、薄宣纸、瓦楞纸、牛皮纸与少量旧印刷纸逐块搭建立面、路面、屋檐、植物、天空和人物。所有主形必须具有清楚可见的露纤维撕边、轻微搭接、纸张厚度、翘起边角和短而自然的实体投影。纸边可以不规则，但街巷外轮廓、道路方向、建筑边界和主体身份必须清晰。禁止光滑矢量切片、激光切割边缘和平面的数码阴影。

配色完全从上方照片提取，将复杂颜色归纳为五至七种灰化纸色，以墙体、路面、屋顶、植物和人物的真实主色为骨架，只保留一处锈红、钴蓝、橄榄绿或其他来自原图的特征色作为强调。不得为所有地点套用固定日本旅游配色。

在连续裸纸区域加入一组铅笔题签。根据照片中可确认的地点、道路结构、光线、天气或动作提炼一个1–3个英文单词的标题；加入字段“PAPER STUDY 01”，多张照片依次编号；再加入一行5–12个英文单词的观察句。标题在正常聊天预览中清楚可读，字段和观察句较小但不能缩成灰线。文字像细石墨或炭笔直接写在纸上，可配一处箭头，不使用光滑数字字体。无法确认地点时使用场景主题，不编造地名、日期或坐标。

整体应像建筑观察、材料实验和旅行出版物之间的一页纸本作品：街巷一眼可辨，纸层真实，边界清楚，信息经过删减，并有足够留白呼吸。

避免手账九宫格、矩形照片拼贴、同构图纸张滤镜、整片背景铺满、儿童剪贴画、通用旅游图标、虚构地标、作者名、账号、Logo、二维码、水印、假中文和乱码。