86: Chapter 85: The Core 86: Chapter 85: The Core The office returned to tranquility, like a study room, where occasionally someone would flaunt a display of affection, seemingly drawing everyone’s attention, but once the incident passed, it would generally be forgotten.
The work in front of everyone was, at that moment, the most important thing.
Jiang Yuan too returned to his seat and continued to select fingerprints that were suitable for him.
He specifically looked for more obscure fingerprints, those with less than ideal conditions, for comparison.
Compared to standard forensic analysis, he was also able to utilize “Image Enhancement Technology (Level 3)” to process the fingerprint images.
In fact, the current automatic fingerprint identification system also processes images.
Refinement, de-noising, ridge separation, and ridge extraction—all of these operations are automatically completed in the background by the computer, sparing the forensic technicians from too much concern.
But given the current level of technology, an automatic fingerprint identification system cannot outperform the manual skills of an expert.
Jiang Yuan directly pulled up a surveillance-like posture and selected an image that was as blurry as the bottom of a pot in just a moment’s time.
Jiang Yuan’s computer tower thus began to stir.
The tower Jiang Yuan was now using was fairly well-equipped, being a new computer centrally purchased by the provincial department.
However, after it was assigned to the forensic workstations, it had practically never exerted its graphics card and CPU for over two years.
At Jiang Yuan’s command, its fans whirred furiously, the noise almost crying out: Why am I so hot?
Jiang Yuan stood up, waved a pack of Zhonghua cigarettes at Li Zemin next to him, and invited him, “How about a cigarette outside?”
Li Zemin found Jiang Yuan’s machine noisy too, and keen to bum a smoke, he immediately followed Jiang Yuan’s steps.
Stepping out of the office, they exchanged a cigarette and began puffing smoke into the air.
Jiang Yuan only took mouthfuls of his cigarette, mostly holding it with his hand while staring blankly ahead.
Li Zemin didn’t care much, only after finishing one cigarette did he let out a satisfied sigh, lit another, and asked, “Why is your computer so loud?”
Jiang Yuan replied, “I’m doing image processing.
Li Zemin knew a bit about it and looked at Jiang Yuan, saying, “Processing fingerprints that loud, your skills must be quite impressive.”
“Yeah, I can do Image Enhancement like the surveillance videos, trying it on fingerprint images to see if it works, but I’m not yet sure about the results.”
“If it works, they should set up a separate space for you, or find a place specifically for you to process images,” Li Zemin remarked with emotion, “You young people are really impressive, unlike us, who struggled to death learning about fingerprints back in the day…
By the way, to what extent can your technology enhance images?”
“It’s hard to say, the blurry ones will become a bit clearer, but it’s targeted,” Jiang Yuan did not elaborate much, as he had not truly experienced it himself.
In general, using Image Enhancement technology on fingerprints is a bit like using a sledgehammer to crack a nut.
Whether it’s suitable or not, he himself couldn’t be sure.
However, one thing was certain.
During the period of the fingerprint competition, the dozens of fingerprint experts on site were, in reality, in competition with each other.
When coming across a fingerprint, the first thing they considered was whether other fingerprint experts had seen it and why it hadn’t been matched.
What was the reason for its non-matching?
If the reason for non-matching was an issue with pairing, then even the best feature points one contributed would be useless.
If the fingerprint was too degraded, it meant that if others couldn’t match it, it was very likely that one couldn’t either.
It was for this reason that some fingerprint experts proposed a strategy for the fingerprint competition: “Don’t work on the clear ones, don’t work on the too vague ones, just try the ones in between.”
It sounded simple, as if it was a strategy one could spin around oneself, but with careful consideration, it was a highly valuable strategy.
In such circumstances, working on very clear fingerprints was, in fact, not of value.
The same was true for the too vague fingerprints.
Only those of moderate clarity were the focus of the work of many fingerprint experts.
However, considering how many fingerprint experts were eyeing these moderately clear fingerprints, that focus also became a difficulty.
Jiang Yuan’s ability to process more blurred images thus became uniquely his own.
Back in the office, Jiang Yuan’s computer had stopped roaring.
The fan was turning weakly, like a special trader who had just entertained a crowd, breathing heavily just to prove it could still breathe, while taking pride in the substantial results fetched from its recent work.
Jiang Yuan looked at the fingerprint image on the screen, noticing many more distinguishable ridges than before.
Jiang Yuan nodded in satisfaction.
Images weren’t feared for being blurred, the fear was having no pixels to work with, all needing to be filled in, making truth and deception indistinguishable in forensic imaging.
An image as blurry as the bottom of a pot was instead simpler; processing the pixels that could be processed, with the rest being usable.
Jiang Yuan tried to choose the more accurate parts of the image to support, then identify the feature points, and finally run the software.
The first time, 100 fingerprints, no match.
The second time, Jiang Yuan selected new feature points, refreshed with 200 fingerprints, and on the 47th one found a similar fingerprint.
Similar, not identical, because both fingerprints had been processed, and there weren’t many places where the feature points matched.
But such situations were most common with challenging fingerprints.
For the same reasons, if the feature points could easily match, why would they be sent to the provincial department, and enter the fingerprint competition?
Here, the computer’s automated judgments were relegated to a secondary role.
The core competitiveness of fingerprint experts always lay in human judgment.
The current fingerprint automatic identification system was, in essence, a numerical system.
It did not make image judgments but instead assigned values to images to make numerical assessments.
Frankly speaking, the fingerprint automatic identification system recognized endpoints and bifurcations; it was effective for most feature points, but for some traditional features, like dots, hooks, crosses, bridges, islands, broken lines, etc., the system lacked discernment.
Furthermore, aspects like the flow direction of the ridge lines or the pattern type, which are not easily articulated, posed even more of a challenge for the system.
In the environment of the fingerprint competition, Jiang Yuan didn’t have to dwell on the computer’s feature points; he annotated according to the traditional Qingdao Single-Finger Fingerprint Analysis Method, then right-clicked to select “Confirm as Identical.”
In the office, the prompt for a new message rang out in unison.
Li Zemin instinctively looked toward Jiang Yuan next to him, only to see Jiang Yuan slowly opening a new pack of Zhonghua.