Ricoh Company,Ltd. Close
Front Runner
Research Theme: Developing Products that Entertain :
   Haike Guan(Ph.D. in engineering)  
It is our customers who find the value in our technology

When I was a student, I researched aspheric surface topology measuring techniques. I studied holography later in Graduate school. I then looked for a place I could use my expertise of image processing technology and selected Ricoh. I was also impressed that Ricoh had high technical capabilities in the optical field, backed up by time-honored results.

My first work after joining the company was "Physical Document Security." The goal was to protect physical documents from security threats using embedded pattern printing technology. Various manufacturers were working on security in computers or networks at the time, but guarding the security of paper output was an idea spawned by Ricoh alone, as the company had already advocated the "integration of paper and electronic documents."

Through my job, I embraced the significance of developing a technology that helps the people of the world. Physical document security technology was installed on Ricoh multi-function printer (MFP) released in 2003. It worked as an "illegal copy guard function." Researchers tend to concentrate only on technical issues and sometimes pass over how society is affected. A manufacturer's R&D should never be merely autotelic research; research is not meaningful if it is not applied to the real world and if it is not evaluated by the market and customers.

R&D contributing to commercialization

Photo:Haike GuanMultiple development themes started in 2004, adding to the image security technology. Themes include image processing technologies for digital cameras called aberration correction of lens (*1), long exposure noise reduction (*2), and skew correction. Skew correction rectifies distortion that results when a square object is filmed diagonally. The first two were based on requests from the digital camera division (Personal Multi-Media Company: PMMC), while skew correction was a function we in the R&D Division proposed. We demonstrated it to PMMC, who enthusiastically said "This is interesting, we’ll put it in our cameras." That’s when development toward commercialization started.

*1 Aberration correction of lens: The geometric distortion produced by the difference of image formation scale factor between the core and periphery of an image is called distortion and brightness reduction of the lens periphery produced by the difference in angle of incident light to an optic axis is called optical vignetting. These distortions, which cannot be suppressed in optical design, can be rectified by image processing technology.

*2 Long exposure noise reduction: The signal emitted by an imaging element (CCD: Charge Coupled Device) appears as image noise in long exposures. The image processing technology to eliminate this is called long exposure noise reduction.

I was appointed theme leader to develop distortion correction technology. This was a big change: I went from working directly on a given technical subject to being responsible for planning the deployment of resources such as personnel, material and money towards achieving the final goal. Although the job carried heavy responsibilities, and I had only a little product development experience, thanks to my boss and team members, I was able to achieve commercialization.

I focused mainly on developing a skew correction algorithm and the software for skew correction. CPU and memory resources (*3) on a digital camera are limited. This is true not only for skew correction but for the distortion correction mentioned earlier. Working to correct distortion and skew at high speed within these limited resources gives developers a chance to show their skills. Development, however, can never be done alone. I broke through each barrier one at a time, cooperating with members of the development group and people in related departments. This technology was installed in the Caplio R3 (released in November, 2005) and is an integral part of Ricoh's digital cameras today.

*3 Resources: Compared to a personal computer, CPU capacity on a general-use digital camera is about 1/100 and memory capacity is about 1/1000.

Bridging the product division and the Beijing research center

SRCB members
Then, development of image processing technology for digital cameras becomes the focus of my work. Around the same time, the company decided to establish Ricoh Software Research Centre Beijing (SRCB). It is the foothold for promoting leading-edge technology development in China, where growth is significant and many high-skilled personnel are available. I worked from start-up through settlement of the digital camera image processing development team. Although all SRCB members had doctorates in engineering, none were experienced in product development. We started by emphasizing the difference between building a prototype in a laboratory and developing a product our customers want to buy.

The first theme at SRCB was to develop face detection technology. It is a function that was already installed in various manufacturers' digital cameras, so Ricoh was the third in the industry to use it. As development started, we felt frustrated about not being the first to market. That, however, drove our team to catch up fast.

While examining how the needs of users would be converted to tangible forms in the specifications with PMMC members, I tried to discover, specifically, how to develop the technology and define the subjects with SRCB people. In guiding SRCB people, who were inexperienced in commercialization, my task was always to define the target clearly. With this approach, when failure became an issue or when we hit a plateau, the clear target dictated what should be given top priority.

Although we started literally from zero, because we were able to use team capabilities to the full, in both Japan and China, we were able to have the face detection function installed in the Caplio R6 in March 2007, a year after development began. PMMC, SRCB members and I all gained valuable experience.

Users' voices reach us directly

Photo:Haike GuanAlthough we played catch up with face detection, the themes that followed were world-first Ricoh technologies. They are multi-pattern auto white balance, wide dynamic range technology, and multi target autofocus technology. We took on these themes again through a coalition of PMMC, SRCB, and the R&D group I belong to. Ricoh expanded its thrust in the digital camera field following the hit made by our GR and Caplio series in the market, which resulted in my being assigned to technology development full time.

We conducted market surveys before loading unexplored new technologies. We thoroughly investigated the kind of functions expected to deliver convenience and pleasure in taking pictures. And we pared down the functions to load. As I said earlier, it is our customers who find the value in our technology, I feel strongly that, with consumer products, the way customers evaluate our technology is transmitted to us directly. Therefore, we looked at user evaluations in the development stage (*4) much more carefully than in the past. Although all were first trials for me, the realization that we were developing world-first functions before anyone else, with our own hands, gave me strength to overcome many difficulties.

*4 User evaluations in development stage: For example, with the multiple target autofocus technology, we started by letting the photographer select the focusing point. However, while proceeding with the user evaluation with the first prototype, we finally adopted a way for the camera to select the focusing point automatically, for convenience. This new Ricoh technology was born through a process of hypothesis and verification.

The new technology was installed on the CX1, which debuted in February 2009, and gained a high reputation in the market. I can't forget even now the joy I felt when reading a user’s blog: "In the past, I couldn’t take a good portrait in the shade, but I became able to take wonderful photographs with this multi pattern auto white balance function."

Dreams come true only with Ricoh

I believe that digital camera technology is still in the development stage. We should make a camera, which lets our users to take a photo they expected easily. At the same time, in addition to being satisfied with the photographs, the photographer will see them as joyful reminders of the moment they were taken. As digital cameras grow in popularity, the shot number of photos will become overwhelming. You may have a problem selecting the best shot from so many. Although still just an idea, what would you think of a function that automatically extracts images that satisfy specific conditions or themes, and then plays them back in a slide show?

Ricoh’s corporate culture encourages us to propose themes we want to pursue and allows us to work on them ourselves. Beyond that, a progressive spirit prevails and we can aggressively take on new technologies. The possibilities are infinite, limited only by individual ambition. I'm presently thinking of a way to build a common platform that allows users to enjoy viewing photographs taken at will and in various ways. Putting it into a single product may be difficult, but I think it may become feasible if various devices and systems are organically linked. I believe this to be a concept I can bring to fruition only because I’m at Ricoh, a company with imaging devices and the solutions to implement them at full capability.

Photo:Haike Guan Haike Guan(Ph.D. in engineering),
Specialty: Physical information engineering in Interdisciplinary Graduate School of Science and Engineering. Completed doctoral course with a Doctor of Engineering degree. Joined Ricoh in 2001. Played broad roles, including R&D promotion in SRCB while working on R&D of image security technology and digital camera image processing technology. Likes to watch baseball games and enjoys swimming and tennis on holidays.
(2009.12)

To Top