If you have a smartphone with a camera, you may have benefited from Lei Zhang’s work without even knowing it. At Hong Kong Polytechnic University, Zhang works on improving image processing and
recognition — research aimed at enhancing resolution and picture quality, removing blur and making color more vivid.
Image recognition makes it possible for computers to enhance the quality of photographs by recognizing the intended image and removing the distortions. Image classification also works to enhance computer recognition of images, but in this case to help computers – even those in digital cameras and smartphones – track and classify images.
The end result of this is better-quality pictures at a lower price. In recent years, camera prices have been dropping, and phone-camera quality improving, thanks in part to these types of innovations. For instance, face recognition on smartphones helps the camera ensure that faces aren’t blurry.
“I like to say that for this kind of research, impact is reflected in the factory,” says Zhang. “Companies care about research, because it’s potentially put in digital cameras to enhance images.”
Zhang notes that In the future, image enhancement and analysis might be used in increasingly diverse digital imaging products. People are using more and more digital products, and these innovations can help with virtually all of them. In the near future, smartphones will have intelligent cameras that can cosmetically detect unique or intriguing objects, and tell users what objects are in any given scene.
High-quality original research
The key to success in research, says Zhang, is high-quality, original papers. Much like academics, industry professionals citing papers are looking for useful and novel ideas. Producing such research can be frustrating and time consuming, but it’s only with this effort that researchers can attract citations.
“As a postdoc, I developed a program and found that the result was very good, much better than others. I emailed my supervisor, but later found out that the program was wrong,” says Zhang. “So it’s very frustrating. You have to check carefully, and I use this as an example to let my students know that being a researcher is a very tough job.”
Research takes lots of time and meticulous effort, in addition to broad thinking and innovative ideas. Technology is an area that moves incredibly fast, but research doesn’t always move at the same speed. Zhang’s solution to this dilemma is for researchers to decide on interesting problems to focus on early in their career. The selection of a fresh research topic is crucial, and that means choosing one that other people will find interesting.
Zhang says that having a quality topic is the most important factor in high-impact research. Sometimes people might find themselves with unproductive research, and in that case re-examining or reworking their focus may be the best way to improve their impact.
“Quality is very important,” says Zhang. “So if you fix the topic and make some very interesting investigation, that’s very important.”
Zhang, though, is overwhelmingly excited and optimistic about future researchers in his field, saying he is often surprised by their ideas and intelligence. Technological research is a whole different field from what it was 10 years ago, but he says the young researchers he encounters are well adapted to the task.