A 1972 magazine centerfold became the standard test image for image processing
In 1973, researchers at the University of Southern California needed a high-quality photo to test image compression software. They cropped a 512-by-512 pixel square from the November 1972 issue of Playboy featuring model Lena Forsén. Known as "Lenna," the image became the industry standard test picture across computer vision research for over four decades due to its mix of detail, shading, and texture.
The Search for a New Benchmark
In June or July of 1973, Alexander Sawchuk, an assistant professor of electrical engineering at the University of Southern California's Signal and Image Processing Institute (SIPI), was working on a conference paper with a graduate student and the lab manager. The team needed a high-quality, full-color photographic image to test their latest image compression algorithms. Standard engineering test images of the era—such as television test charts, high-contrast monochrome geometric patterns, and drab stock photographs—failed to represent the complex, nuanced visual information typical of real-world photography.
A colleague walked into the laboratory holding the November 1972 issue of Playboy magazine, which featured Swedish model Lena Forsén (credited as Lena Söderberg). Looking for an appealing image with rich tonal variation, the researchers tore off the top portion of the magazine centerfold and wrapped it around the drum of their Muirhead wirephoto scanner. The scanner, equipped with analog-to-digital converters for red, green, and blue color channels, digitized a 512-by-512 pixel square at a resolution of 100 lines per inch. This 512-line format matched the memory architecture and processing limits of their computers, creating what would soon become the most famous test file in the history of digital imaging.
Why the Image Worked Technically
The cropped portrait became an instant technical favorite because it contained an unusually complete spectrum of visual challenges within a single frame. Digital image processing algorithms—such as compression, denoising, sharpening, and edge detection—must handle both sharp transitions and gradual gradients. The Lenna image provided both in abundance. The feather boa on her hat offered high-frequency detail, fine textures, and random patterns, while her hair and the hat brim presented sharp, distinct contours that tested edge-detection routines.
At the same time, the image featured broad, flat areas with gentle tonal transitions, particularly across the model's shoulders, face, and the out-of-focus background. These smooth regions made it immediately obvious if a compression algorithm introduced banding, blocking artifacts, or color distortion. Because the human visual system is biologically tuned to recognize subtle flaws in human faces, researchers could instantly judge the perceptual quality of an algorithm's output without relying solely on mathematical error metrics.
Viral Spread Through Early Networks
The USC SIPI lab maintained one of the earliest digital image databases and frequently distributed magnetic tapes containing digitized images to other research institutions across the world. Because the field of digital image processing was young and high-resolution scanning equipment was exceptionally expensive and rare, researchers at other universities and corporate labs gladly adopted USC's library rather than digitizing their own files.
As image processing expanded through the 1970s and 1980s, standardizing on a single shared image allowed researchers to compare their results directly against published findings in peer-reviewed journals. If a new compression scheme produced a higher peak signal-to-noise ratio or a sharper visual output on the Lenna file than a competing method, readers could immediately gauge the improvement. Over four decades, thousands of papers on wavelet transforms, JPEG development, watermarking, and filtering relied on the same 512-pixel crop.
Copyright and Playboy's Stance
For years, the image circulated through the academic community without the formal consent of Playboy Enterprises, which owned the copyright to the original centerfold. By the time the magazine's legal department noticed the widespread use of the photograph in engineering textbooks, scientific journals, and software documentation in the early 1990s, the picture was already an entrenched industry standard.
Rather than issuing cease-and-desist orders or suing academic institutions, Playboy opted for a pragmatic compromise. The publisher decided to overlook the infringement and grant permission for educational and non-commercial scientific research. The company recognized that the image had achieved an unprecedented cultural status within computer science, famously noting that the widespread citation of the issue was an unexpected point of pride.
The Model and the Tech Community
Lena Forsén herself was unaware of her status in computer science for over two decades. After her modeling work in the United States, she returned to Sweden to work in government administration. She only learned about her digital ubiquity in 1988 when she was contacted by a Swedish computer magazine.
In 1997, Forsén was invited as a guest of honor to the 50th Annual Conference of the Society for Imaging Science and Technology (IS&T) in Boston. Attendees formed long lines to meet her and request autographs on printouts of the image they had analyzed for years. While she expressed amusement at her unexpected role in technological history, she later voiced support for retiring the image in favor of modern, diverse benchmarks that reflect contemporary computing environments.
The Push for Retirement and Modern Standards
Beginning in the late 1990s and accelerating through the 2010s, researchers and educators began questioning the continued use of an image sourced from an adult magazine. Critics pointed out that relying on a centerfold as the universal standard created an unwelcoming atmosphere for women in computer science and electrical engineering, reinforcing historical gender imbalances in technical fields.
In response to these concerns, academic journals, professional organizations, and educational platforms began phasing out the image. Publications such as Nature journals and institutions like the IEEE (Institute of Electrical and Electronics Engineers) established policies discouraging or outright banning the submission of new papers that use the Lenna test image. Modern computer vision research now relies on vast, openly licensed datasets containing millions of images, rendering single-image benchmarks like Lenna obsolete.
Key takeaways
•The 'Lenna' image was created in 1973 when researchers at USC digitized a 512x512 pixel crop from a 1972 Playboy centerfold using a drum scanner.
•The picture became ubiquitous because its combination of fine textures (feathers, hair), smooth gradients (skin), and human facial features made it ideal for testing compression and filtering.
•Playboy chose not to enforce copyright claims against academic and research uses after discovering the image's widespread scientific adoption.
•Leading academic bodies and journals, including the IEEE, have banned or discouraged the image in favor of modern, inclusive datasets.